From peter.st.john at gmail.com Mon Nov 1 01:19:58 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Mon, 1 Nov 2010 01:19:58 -0400 Subject: [Beowulf] "Go" programming language Message-ID: Has anyone tried the "Go" programming language on a beowulf? The language's homepage says, " Its concurrency mechanisms make it easy to write programs that get the most out of multicore and networked machines..." (from http://golang.org/) The wiki is http://en.wikipedia.org/wiki/Go_%28programming_language%29 I'm sure I'll use MPI but Google hired some pretty cool language designers. Peter P.S. described somewhere as "merging C++ with Python" which maybe explains an odd white-space rule (open curly bracket can't begin a line because it would confuse automated semicolon line endings), Yuk. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From deadline at eadline.org Mon Nov 1 09:36:48 2010 From: deadline at eadline.org (Douglas Eadline) Date: Mon, 1 Nov 2010 09:36:48 -0400 (EDT) Subject: [Beowulf] SC10 Beowulf Bash Short Interview In-Reply-To: <43163.192.168.93.213.1288401573.squirrel@mail.eadline.org> References: <68A57CCFD4005646957BD2D18E60667B121AD077@milexchmb1.mil.tagmclarengroup.com> <43163.192.168.93.213.1288401573.squirrel@mail.eadline.org> Message-ID: <34986.192.168.93.213.1288618608.squirrel@mail.eadline.org> For those that may be wondering, there is some background on the Beowulf Bash at: http://insidehpc.com/2010/11/01/monday-beowulf-bash-is-the-big-easy-at-sc10/ -- Doug > > It is that time of the year. If you are attending SC10, > here is what you have been waiting for: > > The Big Wheels Keep On Turning Beowulf Bash > > http://www.xandmarketing.com/beobash10/ > > -- > Doug > > -- > This message has been scanned for viruses and > dangerous content by MailScanner, and is > believed to be clean. > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- Doug -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From Bill.Rankin at sas.com Mon Nov 1 11:21:36 2010 From: Bill.Rankin at sas.com (Bill Rankin) Date: Mon, 1 Nov 2010 15:21:36 +0000 Subject: [Beowulf] RE: Storage - the end of RAID? In-Reply-To: <20101029191505.GA29737@bx9.net> References: <68A57CCFD4005646957BD2D18E60667B121AD077@milexchmb1.mil.tagmclarengroup.com> <4CCB168B.6000002@runnersroll.com> <20101029191505.GA29737@bx9.net> Message-ID: <76097BB0C025054786EFAB631C4A2E3C0949E2E8@MERCMBX03R.na.SAS.com> > Um, it's not really RAID 1 when the drives are in different servers. > Although there's not much point in arguing about that. > > -- greg My knee-jerk reaction to Greg's statement was going to be something snotty and along the lines of "what part of 'R' don't you understand?" ;-) But upon further pondering and also going back and re-reading the original article a little more slowly, I think that there is a point in there. Now, one thing I do have an issue with is the article's title claim that RAID is somehow dead. This is clarified in the very first sentence where he identifies that this refers to "costly RAID controllers". Terribly misleading but as someone mentioned earlier it does grab your attention. But in a sense this is old news to us. What the article essentially addresses is that with the huge increase in the I/O capability of other pieces of the system (CPUs, busses, etc) the model of having all your data accessed by going out on the wire and pulling it in from some remote (possibly multiple) high end storage servers cannot survive. At least not for "active" data. Well, duh. Personally I've been using clusters to crush some decent high-end storage arrays since around 2003. Back then the general rule was that we had local disk on the nodes and you would do any significant I/O to those disks. We would stage static copies of input data (eg. Genomics databases) to those disks also just to avoid going out on the wire. So we threw out RAID years ago for our active data. Where we did keep it was for our less active data - home directories, executables, etc - where we did have a need for a definitive, coherent storage image and availability was more important that absolute performance. But this too is starting to change as the node count rises. I think that the current thinking is that while disks have gotten very large, their I/O performance has not kept pace. With the cost of solid state storage coming down in price now, it makes a lot of sense to start replacing disks where we have single point bottlenecks in our I/O chain. So I look at the whole discussion as the realization that finally the rest of the world is catching up to us. :-) -bill _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rtomek at ceti.pl Mon Nov 1 12:24:55 2010 From: rtomek at ceti.pl (Tomasz Rola) Date: Mon, 1 Nov 2010 17:24:55 +0100 (CET) Subject: [Beowulf] "Go" programming language In-Reply-To: References: Message-ID: On Mon, 1 Nov 2010, Peter St. John wrote: > Has anyone tried the "Go" programming language on a beowulf? Probably not, I wildly guess. First of all, the language is too new (just one year old). Besides, it looks like they wanted to have a language that didn't look like C, even thou it looked like it. And they succeded... > The language's homepage says, > " Its concurrency mechanisms make it easy to write programs that get the > most out of multicore and networked machines..." Well, myself, maybe I would stand on my hands, clap my feet and yell "ooom! oom!" if they did this to Haskell (maybe there is some library for distributing Haskell, I'm not sure, had no time/need to check). Personally, I don't think I would use Go. At least in the foreseeable future. It doesn't seem it gives anything that Python and C couldn't give, and they are better established as languages. At best, it may be interesting syntactic experiment. If few years from now there is still some rumour about it, then maybe... > (from http://golang.org/) > > The wiki is http://en.wikipedia.org/wiki/Go_%28programming_language%29 > > I'm sure I'll use MPI but Google hired some pretty cool language designers. Yes. There is a proverb in Polish, that roughly translates as "when you hire six chefs in a kitchen, you have nothing to eat". > Peter > > P.S. described somewhere as "merging C++ with Python" which maybe > explains an odd white-space rule (open curly bracket can't begin a line > because it would confuse automated semicolon line endings), Yuk. Genetically-wise, some hybrids are great and move things forward. But some other are unable to live without life support. Regards, Tomasz Rola -- ** A C programmer asked whether computer had Buddha's nature. ** ** As the answer, master did "rm -rif" on the programmer's home ** ** directory. And then the C programmer became enlightened... ** ** ** ** Tomasz Rola mailto:tomasz_rola at bigfoot.com ** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Mon Nov 1 12:37:36 2010 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Mon, 1 Nov 2010 09:37:36 -0700 Subject: [Beowulf] RE: Storage - the end of RAID? In-Reply-To: <76097BB0C025054786EFAB631C4A2E3C0949E2E8@MERCMBX03R.na.SAS.com> References: <68A57CCFD4005646957BD2D18E60667B121AD077@milexchmb1.mil.tagmclarengroup.com> <4CCB168B.6000002@runnersroll.com> <20101029191505.GA29737@bx9.net> <76097BB0C025054786EFAB631C4A2E3C0949E2E8@MERCMBX03R.na.SAS.com> Message-ID: > I think that the current thinking is that while disks have gotten very large, their I/O performance > has not kept pace. With the cost of solid state storage coming down in price now, it makes a lot of > sense to start replacing disks where we have single point bottlenecks in our I/O chain. Your observation isn't surprising.. over time, density increases (on the disk, on silicon, heck, on punched cards), but the data still has to flow serially through some medium (even if you have a parallel databus, it's not megabits wide), and there, you're limited by EM propagation and all its ills. Speed over a wire, unfortunately, doesn't increase exponentially like density does. It doesn't even get a square law for feature spacing vs areal density. The other issue that bites you pretty hard is power consumption. Fast data = more transitions = more units of charge moving from one place to another = more IR losses. And that's without looking at the more complex transmitting/receiving hardware needed as you move from pushbuttons and relays to things that have to worry about impedance discontinuities and adaptive equalization. In many ways, the whole idea of distributed computing is equally applicable to distributed storage, problems and all, just a matter of the scale whether it's registers in the CPU, cache, some level of RAM, or bulk storage. > > So I look at the whole discussion as the realization that finally the rest of the world is catching up > to us. > In many ways, Beowulfery has helped here.. especially in its early incarnations, the "between node" pipes were pretty slow compared to previous supercomputer designs, so people spent a lot of time figuring out how to structure algorithms so they had good locality of reference and were decoupled at fine time scales. It's sort of the inverse of the classic array processor, systolic array, or even SIMD machines. I *like* having architectures generically based on message passing rather than shared memory. Programming is harder at first, because you need to explicitly recognize the non-deterministic behavior of the messages, but I think it makes the result design cleaner from an architectural standpoint. It really gets rid of the "global shared variable" thing that is a bane of multithreaded programming. (Of course, if you're coming from a tightly coupled environment with fast semaphores, you find it a pain.. ) _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From nixon at nsc.liu.se Tue Nov 2 08:39:27 2010 From: nixon at nsc.liu.se (Leif Nixon) Date: Tue, 02 Nov 2010 13:39:27 +0100 Subject: [Beowulf] Anybody using Redhat HPC Solution in their Beowulf In-Reply-To: <4CC6FD28.1050303@runnersroll.com> (Ellis H. Wilson, III's message of "Tue, 26 Oct 2010 12:09:12 -0400") References: <4CC11EA8.8030602@gmail.com> <68A57CCFD4005646957BD2D18E60667B12154605@milexchmb1.mil.tagmclarengroup.com> <4CC60BFA.4080502@runnersroll.com> <68A57CCFD4005646957BD2D18E60667B12154E23@milexchmb1.mil.tagmclarengroup.com> <4CC6FD28.1050303@runnersroll.com> Message-ID: "Ellis H. Wilson III" writes: > Optimally IMHO, in university setups physical scientists create the > need for HPC. These types shouldn't (as Kilian mentions) need to > inherit all of the responsibilities and overheads of cluster > management to use one (or pay cluster vendors annually for support). > They should simply walk over to the CS department Agh, no, not the CS guys. They should come to their friendly local HPC centre, who will help them spec, acquire and operate a custom cluster (or give them a time allocation on a shared cluster). I'm being purely objective here, of course. -- / Swedish National Infrastructure for Computing Leif Nixon - Security officer < National Supercomputer Centre \ Nordic Data Grid Facility _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From tomislav.maric at gmx.com Tue Nov 2 17:45:51 2010 From: tomislav.maric at gmx.com (tomislav_maric@gmx.com) Date: Tue, 02 Nov 2010 22:45:51 +0100 Subject: [Beowulf] cluster profiling Message-ID: <20101102215301.225020@gmx.com> Hi everyone, I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". Best regards, Tomislav Maric, (MSc Mechanical Engineering, just to clarify my ignorance regarding HPC) _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From hearnsj at googlemail.com Tue Nov 2 19:21:13 2010 From: hearnsj at googlemail.com (John Hearns) Date: Tue, 2 Nov 2010 23:21:13 +0000 Subject: [Beowulf] cluster profiling In-Reply-To: <20101102215301.225020@gmx.com> References: <20101102215301.225020@gmx.com> Message-ID: On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > This sounds very cool. To be honest, most people use Excel spreadsheets to plot this sort of thing. If you can produce an automated framework to do this it would be very interesting. I have to slightly question your choice of Ganglia - have you thought of using sysstat to capture the system's load or memory figures? _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Tue Nov 2 20:49:34 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Tue, 2 Nov 2010 20:49:34 -0400 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: deBeers laser-engraves serial numbers onto their (natural) diamonds (to counter the increasing gem quality of artificial diamonds made by, say, chemical vapor deposition). So how about laser engraving data onto cheap chemical vapor deposition thin diamond slices? (One of the ideas had been to make microelectronics substrates from diamond this way, since diamond conducts heat better than silicon). Peter On Fri, Oct 29, 2010 at 1:36 PM, Robert G. Brown wrote: > On Fri, 29 Oct 2010, Lux, Jim (337C) wrote: > > Or, how about something like the UNICON aka "terabit memory" (TBM) from >> Illiac IV days. It's a stable polyester base with a thin film of rhodium >> that was ablated by a laser making 3 micron holes to write the bits. >> $3.5M >> to store a terabit in 1975. >> > > Burned RO laser disks should in principle be as stable, if the medium > used is thick enough. The problem is that CDs tend to be mass produced > with very thin media, cheap plastic, and are even susceptible to > corrosion through the plastic over time. If one made a CD with tempered > glass and a moderately thick slice of e.g. stainless steel or > platinum... > > But then your problem is the reader. CD readers give way to DVD and are > still backwards compatible, sort of. But what about the 2020 > equivalent? Will there even be one? Nobody will buy actual CDs any > more. Nobody will buy movies on DVDs any more (seriously, I doubt that > they will). Will there BE a laser drive that is backwards compatible to > CD, or will it go the way of reel to reel tapes, 8 track tapes, cassette > tapes, QIC tapes, floppy drives of all flavors (including high capacity > drives like the ones I have carefully saved at home in case I ever need > one), magnetic core memories, large mountable disk packs, exabyte tape > drives, DA tapes, and so on? I rather think it will be gone. It isn't > even clear if hard disk drives will still be available (not that any > computer around would be able to interface with the 5 or 10 MB drives of > my youth anyway). > > This is the problem with electronics. You have to have BOTH long time > scale stability AND an interface for the ages. And the latter is highly > incompatible with e.g. Moore's Law -- not even the humble serial port > has made it through thirty years unscathed. Is the Universal Serial Bus > really Universal? I doubt it. And yet, that is likely to be the only > interface available AT ALL (except for perhaps some sort of wireless > network that isn't even VISIBLE to old peripherals) on the vast bulk of > the machines sold in a mere five years. > > A frightening trend in computing these days is that we may be peaking in > the era where one's computer (properly equipped with a sensible > operating system) is symmetrically capable of functioning as a client > and a server. Desktop computers were clients, servers, or both as one > wished, from the days of Sun workstations through to the present, with > any sort of Unixoid operating system and adequate resources. From the > mid 90's on, with Linux, pure commodity systems were both at the whim of > the system owner -- anybody could add more memory, more disks, a backup > device, and the same chassis was whatever you needed it to be. > > Now, however, this general purpose desktop is all but dead, supplanted > by laptops that are just as powerful, but that lack the expandability > and repurposeability. And laptops are themselves an endangered species > all of a sudden -- in five years a "laptop" could very well be a single > "pad" (touchscreen) of whatever size with or without an external > keyboard, all wireless, smooth as a baby's bottom as far as actual plugs > are concerned (or maybe, just maybe, with a single USB charger/data port > or a couple of slots for SD-of-the-day or USB peripherals). Actual data > storage may well migrate into servers that are completely different > beasts, far away, accessible only over a wireless network, and > controlled by others. > > An enormous step backwards, in other words. A risk to our political > freedom. And yet so seductive, so economical, so convenient, that we > may willingly dance down a primrose path to an information catastrophe > that is more or less impossible still with the vast decentralization of > stored knowledge. > > rgb > > > Robert G. Brown http://www.phy.duke.edu/~rgb/ > Duke University Dept. of Physics, Box 90305 > Durham, N.C. 27708-0305 > Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu > > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From samuel at unimelb.edu.au Tue Nov 2 20:49:59 2010 From: samuel at unimelb.edu.au (Christopher Samuel) Date: Wed, 03 Nov 2010 11:49:59 +1100 Subject: [Beowulf] cluster profiling In-Reply-To: <20101102215301.225020@gmx.com> References: <20101102215301.225020@gmx.com> Message-ID: <4CD0B1B7.9020705@unimelb.edu.au> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 03/11/10 08:45, tomislav_maric at gmx.com wrote: > that will use the Ganglia-python interface and try to > give me an insight into the way machine is burdened > during runs Depending on how old your kernel is the "perf" utility (found in the tools/perf directory in your kernel sources, or packaged in Ubuntu as part of the linux-tools package or linux-tools-2.6 in Debian Squeeze) may well give you some interesting stats. As a here is an overview of stats a "find -ls" over the current kernel git tree: $ perf stat find . -ls > /dev/null Performance counter stats for 'find . -ls': 372.415331 task-clock-msecs # 0.923 CPUs 158 context-switches # 0.000 M/sec 2 CPU-migrations # 0.000 M/sec 395 page-faults # 0.001 M/sec 648855865 cycles # 1742.291 M/sec 698863597 instructions # 1.077 IPC 14321645 cache-references # 38.456 M/sec 379109 cache-misses # 1.018 M/sec 0.403454703 seconds time elapsed You can use the "perf list" command to get a list of all the kernel tracepoints you can monitor and then you can select them individually with the "stat" command. Here is perf monitoring CPU migrations, L1 dcache misses and the kernel scheduler stats of that well known HPC program "top". ;-) perf stat -e migrations -e L1-dcache-load-misses -e sched:* top [...] Performance counter stats for 'top': 0 CPU-migrations # 0.000 M/sec 1038307 L1-dcache-load-misses # 0.000 M/sec 0 sched:sched_kthread_stop # 0.000 M/sec 0 sched:sched_kthread_stop_ret # 0.000 M/sec 0 sched:sched_wait_task # 0.000 M/sec 98 sched:sched_wakeup # 0.000 M/sec 0 sched:sched_wakeup_new # 0.000 M/sec 61 sched:sched_switch # 0.000 M/sec 0 sched:sched_migrate_task # 0.000 M/sec 0 sched:sched_process_free # 0.000 M/sec 1 sched:sched_process_exit # 0.000 M/sec 0 sched:sched_process_wait # 0.000 M/sec 0 sched:sched_process_fork # 0.000 M/sec 15 sched:sched_signal_send # 0.000 M/sec 49 sched:sched_stat_wait # 0.000 M/sec 174 sched:sched_stat_runtime # 0.000 M/sec 67 sched:sched_stat_sleep # 0.000 M/sec 0 sched:sched_stat_iowait # 0.000 M/sec 29.452075124 seconds time elapsed With root access you can even do "perf top" to see what's going on under the hood. You can also use "perf record -g $COMMAND" to record the profiling information for $COMMAND to perf.data along with call graph information so you can display a detailed tree view of what was going on via the "perf report" command. Quite a neat little tool I've got to say! cheers, Chris - -- Christopher Samuel - Senior Systems Administrator VLSCI - Victorian Life Sciences Computational Initiative Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 http://www.vlsci.unimelb.edu.au/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkzQsbcACgkQO2KABBYQAh+6JACaAx7p0zARcGGO4busVv7AbqHL tCcAnA4Z6HOs1LTbucprnyBJFxF6glo+ =D2wX -----END PGP SIGNATURE----- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Tue Nov 2 21:19:22 2010 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Tue, 2 Nov 2010 18:19:22 -0700 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: Diamonds may be almost forever, But I would think that fused silica would work almost as well, and is substantially less expensive. If you want something exotic, how about ion implantation of Cr+ or Ti+ ions into alumina Jim Lux +1(818)354-2075 From: Peter St. John [mailto:peter.st.john at gmail.com] Sent: Tuesday, November 02, 2010 5:50 PM To: Robert G. Brown Cc: Lux, Jim (337C); beowulf at beowulf.org Subject: Re: [Beowulf] Re: Interesting deBeers laser-engraves serial numbers onto their (natural) diamonds (to counter the increasing gem quality of artificial diamonds made by, say, chemical vapor deposition). So how about laser engraving data onto cheap chemical vapor deposition thin diamond slices? (One of the ideas had been to make microelectronics substrates from diamond this way, since diamond conducts heat better than silicon). Peter -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From samuel at unimelb.edu.au Tue Nov 2 22:15:11 2010 From: samuel at unimelb.edu.au (Christopher Samuel) Date: Wed, 03 Nov 2010 13:15:11 +1100 Subject: [Beowulf] "Go" programming language In-Reply-To: References: Message-ID: <4CD0C5AF.5000806@unimelb.edu.au> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 02/11/10 03:24, Tomasz Rola wrote: > Well, myself, maybe I would stand on my hands, clap my feet and yell > "ooom! oom!" if they did this to Haskell (maybe there is some library for > distributing Haskell, I'm not sure, had no time/need to check). One of our guys (Bernie Pope) who is a FP geek has been working on MPI bindings for Haskell in his free time (the older ones had been abandoned some time ago). It was presented at the AusHac Australian Haskell Hackathon this July. Presentation here: http://www.berniepope.id.au/docs/mpi_bindings.pdf Code here: http://github.com/bjpop/haskell-mpi cheers! Chris - -- Christopher Samuel - Senior Systems Administrator VLSCI - Victorian Life Sciences Computational Initiative Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 http://www.vlsci.unimelb.edu.au/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkzQxa4ACgkQO2KABBYQAh8LowCdFbSahWYbdMJR0/Ge9QLKaDxD f8MAn2YdYXuCG/zjnh5pDQBrtdo9BWjg =VXop -----END PGP SIGNATURE----- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rtomek at ceti.pl Tue Nov 2 23:06:30 2010 From: rtomek at ceti.pl (Tomasz Rola) Date: Wed, 3 Nov 2010 04:06:30 +0100 (CET) Subject: [Beowulf] "Go" programming language In-Reply-To: <4CD0C5AF.5000806@unimelb.edu.au> References: <4CD0C5AF.5000806@unimelb.edu.au> Message-ID: On Wed, 3 Nov 2010, Christopher Samuel wrote: > One of our guys (Bernie Pope) who is a FP geek has been working > on MPI bindings for Haskell in his free time (the older ones had > been abandoned some time ago). It was presented at the AusHac > Australian Haskell Hackathon this July. Wow, interesting. Thanks! Regards, Tomasz Rola -- ** A C programmer asked whether computer had Buddha's nature. ** ** As the answer, master did "rm -rif" on the programmer's home ** ** directory. And then the C programmer became enlightened... ** ** ** ** Tomasz Rola mailto:tomasz_rola at bigfoot.com ** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From ntmoore at gmail.com Wed Nov 3 01:16:30 2010 From: ntmoore at gmail.com (Nathan Moore) Date: Wed, 3 Nov 2010 00:16:30 -0500 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: Diamonds will burn, just like coal. Also, at surface of the earth pressure, diamond is (by free-energy calc) the less stable state of carbon, and over (geologic) time diamond will trend to graphite/coal. Schroeder's "Thermal Physics" text talks about this for a few pages. A friend of mine has a wedding band with a tiny "imperfection" in the stone - a streak of graphite that you can see with a hand-lens. It makes the stone less valuable, but frankly its more interesting than your standard hunk of glitter. On Tue, Nov 2, 2010 at 8:19 PM, Lux, Jim (337C) wrote: > Diamonds may be almost forever, > > But I would think that fused silica would work almost as well, and is > substantially less expensive. > > If you want something exotic, how about ion implantation of Cr+ or Ti+ ions > into alumina > > > > Jim Lux > +1(818)354-2075 > > *From:* Peter St. John [mailto:peter.st.john at gmail.com] > *Sent:* Tuesday, November 02, 2010 5:50 PM > *To:* Robert G. Brown > *Cc:* Lux, Jim (337C); beowulf at beowulf.org > > *Subject:* Re: [Beowulf] Re: Interesting > > > > deBeers laser-engraves serial numbers onto their (natural) diamonds (to > counter the increasing gem quality of artificial diamonds made by, say, > chemical vapor deposition). So how about laser engraving data onto cheap > chemical vapor deposition thin diamond slices? (One of the ideas had been to > make microelectronics substrates from diamond this way, since diamond > conducts heat better than silicon). > Peter > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > -- - - - - - - - - - - - - - - - - - - - - - Nathan Moore Associate Professor, Physics Winona State University - - - - - - - - - - - - - - - - - - - - - -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From hearnsj at googlemail.com Wed Nov 3 04:38:10 2010 From: hearnsj at googlemail.com (John Hearns) Date: Wed, 3 Nov 2010 08:38:10 +0000 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: On 29 October 2010 18:36, Robert G. Brown wrote: > ?Will there BE a laser drive that is backwards compatible to > CD, or will it go the way of reel to reel tapes, 8 track tapes, cassette > tapes, QIC tapes, floppy drives of all flavors (including high capacity > drives like the ones I have carefully saved at home in case I ever need > one), magnetic core memories, large mountable disk packs, exabyte tape > drives, DA tapes, and so on? ?I rather think it will be gone. ?It isn't > even clear if hard disk drives will still be available (not that any > computer around would be able to interface with the 5 or 10 MB drives of > my youth anyway). Have a look at the other thread I started, regarding new approaches to data storage. > > Now, however, this general purpose desktop is all but dead, supplanted... ..... > ?Actual data > storage may well migrate into servers that are completely different > beasts, far away, accessible only over a wireless network, and > controlled by others. > > An enormous step backwards, in other words. ?A risk to our political > freedom. ?And yet so seductive, so economical, so convenient, that we > may willingly dance down a primrose path to an information catastrophe > that is more or less impossible still with the vast decentralization of > stored knowledge. I'm not so sure it IS a step backwards. What I think we should be doing is working towards a media-agnostic form of storing data. A recognition that scientific data (and other forms, like movies and music etc.) will carry with them metadata and that the data will migrate through many types of physical media in its lifetime, and will from the outset have multiple copies made. I guess the HPC Grid computing types are doing this already, what I'm rather thinking about is a universal standard for this, and a way of carrying the metadata with the actual data in a way it cannot be lost. Its also funny that I use the term "lifetime" - I guess in the past we all have assumed digital data will have an infinite lifetime, as as discussed above it has come to pass that the decay of media, or reading apparatus being unavailable has made data have a finite lifetime. The real point I am making here is that with cloud type data storage over IP connections even in HPC we will be seeing data accessed not on SCSI volumes (be that direct SCSI, fibrechannel, iSCSI, RAID etc) but from an HTTP accessed object store. You might then say that "Hey - performance matters and that's why we still have SCSI" - I would counter that you will see home users accessing data via ADSL, business users via gigabit, and those HPC class systems will have 10 / 40 / 100 gigabit interfaces. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rgb at phy.duke.edu Wed Nov 3 06:55:41 2010 From: rgb at phy.duke.edu (Robert G. Brown) Date: Wed, 3 Nov 2010 06:55:41 -0400 (EDT) Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: On Wed, 3 Nov 2010, John Hearns wrote: > What I think we should be doing is working towards a media-agnostic > form of storing data. Media agnostic with OPEN codices. A great curse on all such IT mechanisms is the closed codex. ASCII works because it is utterly open, as is its logical extension to UTF. But what about music? What about movies? What about books? What about spreadsheets, processed text documents, etc etc? Perhaps html5 will magically solve all such problems, but I doubt it. > A recognition that scientific data (and other forms, like movies and > music etc.) will carry with them metadata and that the data will > migrate through many types of physical media in its lifetime, and will > from the outset have multiple copies made. > I guess the HPC Grid computing types are doing this already, what I'm > rather thinking about is a universal standard for this, and a way of > carrying the metadata with the actual data in a way it cannot be lost. I think this is all dead on correct, but bearing in mind the forces of darkness arrayed on the other side of this, concerned with everything from DRM to encryption to owning and controlling the codex, I personally am not holding my breath. There are also numerous purely technical issues -- modulus problems, for example, in conversion between ogg and mp3 that result in artifacts when switching between lossy compression algorithms that result in nonlinear degradation of information. Similar issues when dealing with old VGA vs 1080p and so on. None of which will go away as the technology evolves. I'm not certain that this is a truly solvable problem. > Its also funny that I use the term "lifetime" - I guess in the past > we all have assumed digital data will have an infinite lifetime, as as > discussed above it has come to pass that the decay of media, or > reading apparatus being unavailable has made data have a finite > lifetime. > > The real point I am making here is that with cloud type data storage > over IP connections even in HPC we will be seeing data accessed not on > SCSI volumes (be that direct SCSI, fibrechannel, iSCSI, RAID etc) but > from an HTTP accessed object store. You might then say that "Hey - > performance matters and that's why we still have SCSI" - I would > counter that you will see home users accessing data via ADSL, business > users via gigabit, and those HPC class systems will have 10 / 40 / 100 > gigabit interfaces. All of which is groovy and I would never argue, but that doesn't address the relative vulnerability of centralized data both to certain kinds of attack and to other kinds of accidents. Or to political control. If Google (ultimately) controls all the data, who controls Google? What happens if they use it for evil instead of for good? How could one stop them from using it for evil if they have your data and also provide you with all of the software you are using to access that data? Who will, after all, guard the guardians? rgb > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From eugen at leitl.org Wed Nov 3 09:37:02 2010 From: eugen at leitl.org (Eugen Leitl) Date: Wed, 3 Nov 2010 14:37:02 +0100 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: <20101103133701.GH28998@leitl.org> On Tue, Nov 02, 2010 at 08:49:34PM -0400, Peter St. John wrote: > deBeers laser-engraves serial numbers onto their (natural) diamonds (to > counter the increasing gem quality of artificial diamonds made by, say, > chemical vapor deposition). So how about laser engraving data onto cheap > chemical vapor deposition thin diamond slices? (One of the ideas had been to > make microelectronics substrates from diamond this way, since diamond > conducts heat better than silicon). There's a variant of microfiches which uses photoglass. Areas exposed to light produce crystal nuclei. The glass is heated close to Tg at which stage those nuclei initiate crystal growth. Should have lots longer shelf half life than mere microfiches which are good for about century or more under optimal storage conditions. -- Eugen* Leitl leitl http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From tomislav.maric at gmx.com Wed Nov 3 10:06:27 2010 From: tomislav.maric at gmx.com (tomislav_maric@gmx.com) Date: Wed, 03 Nov 2010 15:06:27 +0100 Subject: [Beowulf] Re: cluster profiling (John Hearns) & (Christopher Samuel) Message-ID: <20101103142415.14270@gmx.com> Thanks a lot for the advice. :) I don't want to overcomplicate things. Its just that I really haven't found any literature that explains how to profile in the top level the cluster and decide where the bottlenecks are. The Ganglia monitoring system has a gstat utility and a python interface that returns the monitoring statistics in a raw mode (not in the HTTP form for its web interface). With the ability to use its Python interface I can write the profiling application within a single language / environment, so this is why I chose it, and the simple way to get the info for each node. I've read online that it has some overhead, but I'm dealing with dedicated multicore nodes and my simulations are really computationally intensive, which will phase out the system side processes on the nodes. I will look up the perf and sysstat definitely in detail before coding anything. Please don't get me wrong, I just want to find a way to build a smal ~20 node multicore COTS beowulf for use with OpenFOAM. This would more than suffice for my needs. In order to do this I don't want to overcomplicate things... I just want to learn as much as I can and build my machine. This kind of profiling harness should inform me on the major metrics for the nodes and for the frontend. I have spent significant amount of time on profiling the application (a solver of the OpenFOAM package) on a single system, only to gradually realize that there are multiple dozens of different CFD/CCM solvers in OpenFOAM, that involve different numerical schemes (spatial, temporal) and operate on different physical fields. This is an impossible task, to optimise the ordinary PC (CPU clock rate, RAM amount, etc) for OpenFOAM, because the number of the parameters involves huge number of permutations. The optimisation itself would take too long and the conclusion is still untouched by the reality: I should just buy the before-last family of processors and RAM, and stack up on the machines. Now, I would just like a few pointers in the right directions regarding the macroscopic metrics of the networked cluster: e.g. increase the simulation size while keeping the core number and the node number constant and find out when the switch drops dead. This will help me evaluate for a specific set of cases (up to 3 Million cells), how much nodes do I need to buy, do I need 10 gig eth, and what kind of speedup did the separation of the network into DATA - HSI - FRONTEND bring me. This also reduces the number of parameters to 1 - node number (in this case 1 or 2, but I'm getting more machines soon) 2 - core number (2 - 8, more to come) 3 - mesh density (scripted increase from 200k to 1 M) the rest are the global metrics: used memory, CPU %, duration of the simulation (user CPU time), network traffic (MB/s) on the switch. The metrics stays the same and is temporaly averaged over the course of the simulation (the case decomposition is static, otherwise, no conclusions may be drawn). I have read the book from Robert Lucke, Robert Gordon Brown, HPC for dummies, the Beowulf books, and a swarm of articles online... and well.... I did my homework... still, I could really use any advice anyone can spare on the profiling/scaling of such machine. Thanks again, Tomislav > ----- Original Message ----- > From: beowulf-request at beowulf.org > Sent: 11/03/10 02:22 AM > To: beowulf at beowulf.org > Subject: Beowulf Digest, Vol 81, Issue 3 > > Send Beowulf mailing list submissions to > beowulf at beowulf.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://www.beowulf.org/mailman/listinfo/beowulf > or, via email, send a message with subject or body 'help' to > beowulf-request at beowulf.org > > You can reach the person managing the list at > beowulf-owner at beowulf.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Beowulf digest..." > > > Today's Topics: > > ?1. cluster profiling (tomislav_maric at gmx.com) > ?2. Re: cluster profiling (John Hearns) > ?3. Re: Re: Interesting (Peter St. John) > ?4. Re: cluster profiling (Christopher Samuel) > ?5. RE: Re: Interesting (Lux, Jim (337C)) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Tue, 02 Nov 2010 22:45:51 +0100 > From: "tomislav_maric at gmx.com" > Subject: [Beowulf] cluster profiling > To: beowulf at beowulf.org > Message-ID: <20101102215301.225020 at gmx.com> > Content-Type: text/plain; charset="utf-8" > > Hi everyone, > > I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. > > The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". > > Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > > Best regards, > Tomislav Maric, (MSc Mechanical Engineering, just to clarify my ignorance regarding HPC) > > > > ------------------------------ > > Message: 2 > Date: Tue, 2 Nov 2010 23:21:13 +0000 > From: John Hearns > Subject: Re: [Beowulf] cluster profiling > To: Beowulf Mailing List > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > > > > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > > > > This sounds very cool. > To be honest, most people use Excel spreadsheets to plot this sort of thing. > If you can produce an automated framework to do this it would be very > interesting. > > I have to slightly question your choice of Ganglia - have you thought > of using sysstat to capture the system's load or memory figures? > > > ------------------------------ > > Message: 3 > Date: Tue, 2 Nov 2010 20:49:34 -0400 > From: "Peter St. John" > Subject: Re: [Beowulf] Re: Interesting > To: "Robert G. Brown" > Cc: "beowulf at beowulf.org" , "Lux, Jim \(337C\)" > > Message-ID: > > Content-Type: text/plain; charset="iso-8859-1" > > deBeers laser-engraves serial numbers onto their (natural) diamonds (to > counter the increasing gem quality of artificial diamonds made by, say, > chemical vapor deposition). So how about laser engraving data onto cheap > chemical vapor deposition thin diamond slices? (One of the ideas had been to > make microelectronics substrates from diamond this way, since diamond > conducts heat better than silicon). > Peter > > On Fri, Oct 29, 2010 at 1:36 PM, Robert G. Brown wrote: > > > On Fri, 29 Oct 2010, Lux, Jim (337C) wrote: > > > > Or, how about something like the UNICON aka "terabit memory" (TBM) from > >> Illiac IV days. It's a stable polyester base with a thin film of rhodium > >> that was ablated by a laser making 3 micron holes to write the bits. > >> $3.5M > >> to store a terabit in 1975. > >> > > > > Burned RO laser disks should in principle be as stable, if the medium > > used is thick enough. The problem is that CDs tend to be mass produced > > with very thin media, cheap plastic, and are even susceptible to > > corrosion through the plastic over time. If one made a CD with tempered > > glass and a moderately thick slice of e.g. stainless steel or > > platinum... > > > > But then your problem is the reader. CD readers give way to DVD and are > > still backwards compatible, sort of. But what about the 2020 > > equivalent? Will there even be one? Nobody will buy actual CDs any > > more. Nobody will buy movies on DVDs any more (seriously, I doubt that > > they will). Will there BE a laser drive that is backwards compatible to > > CD, or will it go the way of reel to reel tapes, 8 track tapes, cassette > > tapes, QIC tapes, floppy drives of all flavors (including high capacity > > drives like the ones I have carefully saved at home in case I ever need > > one), magnetic core memories, large mountable disk packs, exabyte tape > > drives, DA tapes, and so on? I rather think it will be gone. It isn't > > even clear if hard disk drives will still be available (not that any > > computer around would be able to interface with the 5 or 10 MB drives of > > my youth anyway). > > > > This is the problem with electronics. You have to have BOTH long time > > scale stability AND an interface for the ages. And the latter is highly > > incompatible with e.g. Moore's Law -- not even the humble serial port > > has made it through thirty years unscathed. Is the Universal Serial Bus > > really Universal? I doubt it. And yet, that is likely to be the only > > interface available AT ALL (except for perhaps some sort of wireless > > network that isn't even VISIBLE to old peripherals) on the vast bulk of > > the machines sold in a mere five years. > > > > A frightening trend in computing these days is that we may be peaking in > > the era where one's computer (properly equipped with a sensible > > operating system) is symmetrically capable of functioning as a client > > and a server. Desktop computers were clients, servers, or both as one > > wished, from the days of Sun workstations through to the present, with > > any sort of Unixoid operating system and adequate resources. From the > > mid 90's on, with Linux, pure commodity systems were both at the whim of > > the system owner -- anybody could add more memory, more disks, a backup > > device, and the same chassis was whatever you needed it to be. > > > > Now, however, this general purpose desktop is all but dead, supplanted > > by laptops that are just as powerful, but that lack the expandability > > and repurposeability. And laptops are themselves an endangered species > > all of a sudden -- in five years a "laptop" could very well be a single > > "pad" (touchscreen) of whatever size with or without an external > > keyboard, all wireless, smooth as a baby's bottom as far as actual plugs > > are concerned (or maybe, just maybe, with a single USB charger/data port > > or a couple of slots for SD-of-the-day or USB peripherals). Actual data > > storage may well migrate into servers that are completely different > > beasts, far away, accessible only over a wireless network, and > > controlled by others. > > > > An enormous step backwards, in other words. A risk to our political > > freedom. And yet so seductive, so economical, so convenient, that we > > may willingly dance down a primrose path to an information catastrophe > > that is more or less impossible still with the vast decentralization of > > stored knowledge. > > > > rgb > > > > > > Robert G. Brown http://www.phy.duke.edu/~rgb/ > > Duke University Dept. of Physics, Box 90305 > > Durham, N.C. 27708-0305 > > Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu > > > > > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > > http://www.beowulf.org/mailman/listinfo/beowulf > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://www.scyld.com/pipermail/beowulf/attachments/20101102/09fdea41/attachment-0001.html > > ------------------------------ > > Message: 4 > Date: Wed, 03 Nov 2010 11:49:59 +1100 > From: Christopher Samuel > Subject: Re: [Beowulf] cluster profiling > To: beowulf at beowulf.org > Message-ID: <4CD0B1B7.9020705 at unimelb.edu.au> > Content-Type: text/plain; charset=UTF-8 > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On 03/11/10 08:45, tomislav_maric at gmx.com wrote: > > > that will use the Ganglia-python interface and try to > > give me an insight into the way machine is burdened > > during runs > > Depending on how old your kernel is the "perf" utility > (found in the tools/perf directory in your kernel sources, > or packaged in Ubuntu as part of the linux-tools package > or linux-tools-2.6 in Debian Squeeze) may well give you > some interesting stats. > > As a here is an overview of stats a "find -ls" over > the current kernel git tree: > > $ perf stat find . -ls > /dev/null > > ?Performance counter stats for 'find . -ls': > > ?372.415331 task-clock-msecs # 0.923 CPUs > ?158 context-switches # 0.000 M/sec > ?2 CPU-migrations # 0.000 M/sec > ?395 page-faults # 0.001 M/sec > ?648855865 cycles # 1742.291 M/sec > ?698863597 instructions # 1.077 IPC > ?14321645 cache-references # 38.456 M/sec > ?379109 cache-misses # 1.018 M/sec > > ?0.403454703 seconds time elapsed > > > You can use the "perf list" command to get a list of all > the kernel tracepoints you can monitor and then you can > select them individually with the "stat" command. > > Here is perf monitoring CPU migrations, L1 dcache misses > and the kernel scheduler stats of that well known HPC > program "top". ;-) > > perf stat -e migrations -e L1-dcache-load-misses -e sched:* top > > [...] > > ?Performance counter stats for 'top': > > ?0 CPU-migrations # 0.000 M/sec > ?1038307 L1-dcache-load-misses # 0.000 M/sec > ?0 sched:sched_kthread_stop # 0.000 M/sec > ?0 sched:sched_kthread_stop_ret # 0.000 M/sec > ?0 sched:sched_wait_task # 0.000 M/sec > ?98 sched:sched_wakeup # 0.000 M/sec > ?0 sched:sched_wakeup_new # 0.000 M/sec > ?61 sched:sched_switch # 0.000 M/sec > ?0 sched:sched_migrate_task # 0.000 M/sec > ?0 sched:sched_process_free # 0.000 M/sec > ?1 sched:sched_process_exit # 0.000 M/sec > ?0 sched:sched_process_wait # 0.000 M/sec > ?0 sched:sched_process_fork # 0.000 M/sec > ?15 sched:sched_signal_send # 0.000 M/sec > ?49 sched:sched_stat_wait # 0.000 M/sec > ?174 sched:sched_stat_runtime # 0.000 M/sec > ?67 sched:sched_stat_sleep # 0.000 M/sec > ?0 sched:sched_stat_iowait # 0.000 M/sec > > ?29.452075124 seconds time elapsed > > With root access you can even do "perf top" to see what's > going on under the hood. > > You can also use "perf record -g $COMMAND" to record the profiling > information for $COMMAND to perf.data along with call graph information > so you can display a detailed tree view of what was going on via the > "perf report" command. > > Quite a neat little tool I've got to say! > > cheers, > Chris > - -- > ?Christopher Samuel - Senior Systems Administrator > ?VLSCI - Victorian Life Sciences Computational Initiative > ?Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 > ?http://www.vlsci.unimelb.edu.au/ > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.10 (GNU/Linux) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iEYEARECAAYFAkzQsbcACgkQO2KABBYQAh+6JACaAx7p0zARcGGO4busVv7AbqHL > tCcAnA4Z6HOs1LTbucprnyBJFxF6glo+ > =D2wX > -----END PGP SIGNATURE----- > > > ------------------------------ > > Message: 5 > Date: Tue, 2 Nov 2010 18:19:22 -0700 > From: "Lux, Jim (337C)" > Subject: RE: [Beowulf] Re: Interesting > To: "Peter St. John" > Cc: "beowulf at beowulf.org" > Message-ID: > > > Content-Type: text/plain; charset="us-ascii" > > Diamonds may be almost forever, > But I would think that fused silica would work almost as well, and is substantially less expensive. > If you want something exotic, how about ion implantation of Cr+ or Ti+ ions into alumina > > Jim Lux > +1(818)354-2075 > From: Peter St. John [mailto:peter.st.john at gmail.com] > Sent: Tuesday, November 02, 2010 5:50 PM > To: Robert G. Brown > Cc: Lux, Jim (337C); beowulf at beowulf.org > Subject: Re: [Beowulf] Re: Interesting > > deBeers laser-engraves serial numbers onto their (natural) diamonds (to counter the increasing gem quality of artificial diamonds made by, say, chemical vapor deposition). So how about laser engraving data onto cheap chemical vapor deposition thin diamond slices? (One of the ideas had been to make microelectronics substrates from diamond this way, since diamond conducts heat better than silicon). > Peter > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://www.scyld.com/pipermail/beowulf/attachments/20101102/370656f6/attachment.html > > ------------------------------ > > _______________________________________________ > Beowulf mailing list > Beowulf at beowulf.org > http://www.beowulf.org/mailman/listinfo/beowulf > > > End of Beowulf Digest, Vol 81, Issue 3 > ************************************** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Wed Nov 3 10:49:31 2010 From: prentice at ias.edu (Prentice Bisbal) Date: Wed, 03 Nov 2010 10:49:31 -0400 Subject: [Beowulf] cluster profiling In-Reply-To: <20101102215301.225020@gmx.com> References: <20101102215301.225020@gmx.com> Message-ID: <4CD1767B.8090701@ias.edu> tomislav_maric at gmx.com wrote: > Hi everyone, > > I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. > > The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". > > Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > Have you looked at something like Vampir for MPI profiling? Support for VampirTrace is built into OpenMPI, if you compile Open MPI wih the correct options. The rub is that I think you need to pay for a Vampir GUI to analyze the data. I've never used it myself, but I saw a demo once, and it looked pretty powerful. http://www.vampir.eu/ You might also want to look at Tau, PAPI, and Perfmon2 http://www.cs.uoregon.edu/research/tau/home.php http://icl.cs.utk.edu/papi/ http://perfmon2.sourceforge.net/ I set this up for one of my users a couple of years ago. I could be wrong, but I think Tau requires PAPI, and PAPI in turn requires the perfmon2 kernel patches. I could be wrong, since it's been a couple of years. Reading the docs above should point you in the correct direction. That's probably more than you wanted to know. -- Prentice _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From j.wender at science-computing.de Wed Nov 3 11:36:57 2010 From: j.wender at science-computing.de (Jan Wender) Date: Wed, 03 Nov 2010 16:36:57 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: <4CD1767B.8090701@ias.edu> References: <20101102215301.225020@gmx.com> <4CD1767B.8090701@ias.edu> Message-ID: <4CD18199.2030807@science-computing.de> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi all, On 11/03/2010 03:49 PM, Prentice Bisbal wrote: > http://www.vampir.eu/ for some time now (maybe start of 2010) Vampir is not available on its own any more. It got included in intels Cluster Toolkit and is available from them (and underwent a name change if I remember correctly). Cheerio, Jan Wender - -- Jan Wender j.wender at science-computing.de Senior IT Consultant Vertrieb science + computing ag A Bull Group company Hagellocher Weg 73 phone +49 7071 9457 257 72070 Tuebingen, Germany fax +49 7071 9457 522 www.science-computing.de - ---- Company Information ------------------------------------------------------- Vorstand/Board of Management: Dr. Bernd Finkbeiner, Dr. Roland Niemeier, Dr. Arno Steitz, Dr. Ingrid Zech Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Michel Lepert Sitz/Registered Office: Tuebingen Registergericht/Registration Court: Stuttgart Registernummer/Commercial Register No.: HRB 382196 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkzRgZkACgkQXxf+IhbDzz4K1wCeJHKPUcKXszndDB/qeIl7G+4x rrMAn2fzJVxK3Whq/wd2FGKvkAo0VISm =A9mv -----END PGP SIGNATURE----- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From worringen at googlemail.com Wed Nov 3 13:06:25 2010 From: worringen at googlemail.com (Joachim Worringen) Date: Wed, 3 Nov 2010 18:06:25 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: References: <20101102215301.225020@gmx.com> Message-ID: On Wed, Nov 3, 2010 at 12:21 AM, John Hearns wrote: > On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > >> >> Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". >> > > This sounds very cool. > To be honest, most people use Excel spreadsheets to plot this sort of thing. > If you can produce an automated framework to do this it would be very > interesting. Be sure to check out http://perfbase.tigris.org, which is a Python/PostgreSQL toolkit for exactly this purpose. It saves you the nightmare of managing, analysing and visualizing experiment data this crude, manual way. I use it (and designed, write and maintain it) for 5 years now, currently with application to my job in the financial sector. My database currently has some 400GB of data in it. Release 1.2.0 was just released yesterday, which included parallized queries to speed them up. Joachim _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From worringen at googlemail.com Thu Nov 4 04:37:30 2010 From: worringen at googlemail.com (Joachim Worringen) Date: Thu, 4 Nov 2010 09:37:30 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: References: <20101102215301.225020@gmx.com> Message-ID: On Wed, Nov 3, 2010 at 12:21 AM, John Hearns wrote: > On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > >> >> Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". >> > > This sounds very cool. > To be honest, most people use Excel spreadsheets to plot this sort of thing. > If you can produce an automated framework to do this it would be very > interesting. This is exactly why I wrote "perfbase" some years ago (see http://perfbase.tigris.org), a toolkit for experiment managment and data analysis. I still maintain it and use it on a daily basis (release 1.2.0 came out this week). It parses benchmark output files "automatically", stores data in a database and lets you create and run very elaborate queries on it, generating gnuplots, OpenOffice spreadsheets, raw text or XML output. Joachim _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From tomislav.maric at gmx.com Sat Nov 6 03:37:41 2010 From: tomislav.maric at gmx.com (tomislav_maric@gmx.com) Date: Sat, 06 Nov 2010 08:37:41 +0100 Subject: [Beowulf] Re: cluster profiling (Prentice Bisbal Message-ID: <20101106073741.14260@gmx.com> Thank you very much for the advice. I think that I will try to get the coarse picture of the system load before going into the MPI profiling of the application itself. Since no processes (intensive ones) will run during the testing, I believe it will give me an overview of the system status and let me see where did the bottleneck appear. After this is finished, I will look ino the MPI profiling of the app itself. Thank you again, Tomislav > ----- Original Message ----- > From: beowulf-request at beowulf.org > Sent: 11/03/10 08:00 PM > To: beowulf at beowulf.org > Subject: Beowulf Digest, Vol 81, Issue 5 > > Send Beowulf mailing list submissions to > beowulf at beowulf.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://www.beowulf.org/mailman/listinfo/beowulf > or, via email, send a message with subject or body 'help' to > beowulf-request at beowulf.org > > You can reach the person managing the list at > beowulf-owner at beowulf.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Beowulf digest..." > > > Today's Topics: > > ?1. Re: cluster profiling (Prentice Bisbal) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 03 Nov 2010 10:49:31 -0400 > From: Prentice Bisbal > Subject: Re: [Beowulf] cluster profiling > To: Beowulf Mailing List > Message-ID: <4CD1767B.8090701 at ias.edu> > Content-Type: text/plain; charset=UTF-8 > > > > tomislav_maric at gmx.com wrote: > > Hi everyone, > > > > I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. > > > > The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". > > > > Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". > > > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > > > > Have you looked at something like Vampir for MPI profiling? Support for > VampirTrace is built into OpenMPI, if you compile Open MPI wih the > correct options. > > The rub is that I think you need to pay for a Vampir GUI to analyze the > data. I've never used it myself, but I saw a demo once, and it looked > pretty powerful. > > http://www.vampir.eu/ > > You might also want to look at Tau, PAPI, and Perfmon2 > > http://www.cs.uoregon.edu/research/tau/home.php > http://icl.cs.utk.edu/papi/ > http://perfmon2.sourceforge.net/ > > I set this up for one of my users a couple of years ago. I could be > wrong, but I think Tau requires PAPI, and PAPI in turn requires the > perfmon2 kernel patches. I could be wrong, since it's been a couple of > years. Reading the docs above should point you in the correct direction. > > That's probably more than you wanted to know. > > > > -- > Prentice > > > ------------------------------ > > _______________________________________________ > Beowulf mailing list > Beowulf at beowulf.org > http://www.beowulf.org/mailman/listinfo/beowulf > > > End of Beowulf Digest, Vol 81, Issue 5 > ************************************** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From dnlombar at ichips.intel.com Mon Nov 8 16:34:23 2010 From: dnlombar at ichips.intel.com (David N. Lombard) Date: Mon, 8 Nov 2010 13:34:23 -0800 Subject: [Beowulf] cluster profiling In-Reply-To: <4CD18199.2030807@science-computing.de> References: <20101102215301.225020@gmx.com> <4CD1767B.8090701@ias.edu> <4CD18199.2030807@science-computing.de> Message-ID: <20101108213422.GA5621@nlxcldnl2.cl.intel.com> On Wed, Nov 03, 2010 at 08:36:57AM -0700, Jan Wender wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi all, > > On 11/03/2010 03:49 PM, Prentice Bisbal wrote: > > http://www.vampir.eu/ > > for some time now (maybe start of 2010) Vampir is not available on its own any > more. It got included in intels Cluster Toolkit and is available from them (and > underwent a name change if I remember correctly). The Intel Trace Analyzer and Collector (ITAC), successor of Pallas Vampir, is available in the Intel Cluster Toolkit. But, that's not a recent event, Intel acquired the HPC group of Pallas in 2003. BTW, the Intel MPI Benchmarks, ex Pallas MPI Benchmarks, are similarly available. -- David N. Lombard, Intel, Irvine, CA I do not speak for Intel Corporation; all comments are strictly my own. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Wed Nov 17 14:08:19 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Wed, 17 Nov 2010 14:08:19 -0500 Subject: [Beowulf] LAPACK C/C++ wrappers Message-ID: I'm going to dabble a bit with writing some C/C++ with a linear algebra library. My fortran is rusty (although I did some work translating fortran to C in this millennium :-). The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice have any real bearing on compiling against MPI later? Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) I'd be glad to hear it. I just installed (k)ubuntu, which seems fine for my purposes so far. I do miss fvwm, and the default "single click to invoke an application" is totally nuts but I fixed it. Thanks, Peter -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From ispmarin at gmail.com Sun Nov 21 23:55:02 2010 From: ispmarin at gmail.com (Ivan Marin) Date: Sun, 21 Nov 2010 22:55:02 -0600 Subject: [Beowulf] LAPACK C/C++ wrappers In-Reply-To: References: Message-ID: I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I did some tests in the past for simple stuff, and worked, but very simple linear system solvers. Ivan Marin Civil Engineering Dept University of Minnesota 500 Pillsbury Dr. S.E. Minneapolis, MN 55455 Laborat?rio de Hidr?ulica Computacional - LHC Departamento de Hidr?ulica e Saneamento - SHS Escola de Engenharia de S?o Carlos - EESC Universidade de S?o Paulo - USP http://albatroz.shs.eesc.usp.br +55 16 3373 8270 2010/11/17 Peter St. John : > I'm going to dabble a bit with writing some C/C++ with a linear algebra > library. My fortran is rusty (although I did some work translating fortran > to C in this millennium :-). > The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice have > any real bearing on compiling against MPI later? > Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) I'd > be glad to hear it. > > I just installed (k)ubuntu, which seems fine for my purposes so far. I do > miss fvwm, and the default "single click to invoke an application" is > totally nuts but I fixed it. > Thanks, > Peter > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From j.wender at science-computing.de Mon Nov 22 04:29:33 2010 From: j.wender at science-computing.de (Jan Wender) Date: Mon, 22 Nov 2010 10:29:33 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: <4CD18199.2030807@science-computing.de> References: <20101102215301.225020@gmx.com> <4CD1767B.8090701@ias.edu> <4CD18199.2030807@science-computing.de> Message-ID: <20101122092932.GA28804@sanja.science-computing.de> Hi, I have to correct myself... On Wed, Nov 03, 2010 at 04:36:57PM +0100, Jan Wender wrote: > for some time now (maybe start of 2010) Vampir is not available on its own any > more. It got included in intels Cluster Toolkit and is available from them (and > underwent a name change if I remember correctly). At SC10 I found the Vampir booth and had a short chat with them. Their tool was included at one time with the intel Cluster tools, but they did not stop to develop it further on their own. You can get their version on www.vampir.eu and http://tu-dresden.de/die_tu_dresden/zentrale_einrichtungen/zih/forschung/software_werkzeuge_zur_unterstuetzung_von_programmierung_und_optimierung/ Cheerio, Jan -- Jan Wender j.wender at science-computing.de Senior IT Consultant Vertrieb science + computing ag A Bull Group Company Hagellocher Weg 73 phone +49 7071 9457 257 72070 Tuebingen, Germany fax +49 7071 9457 522 www.science-computing.de ---- Company Information ------------------------------------------------------- Vorstand/Board of Management: Dr. Bernd Finkbeiner, Dr. Roland Niemeier, Dr. Arno Steitz, Dr. Ingrid Zech Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Michel Lepert Sitz/Registered Office: Tuebingen Registergericht/Registration Court: Stuttgart Registernummer/Commercial Register No.: HRB 382196 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: not available URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From eugen at leitl.org Mon Nov 22 10:50:07 2010 From: eugen at leitl.org (Eugen Leitl) Date: Mon, 22 Nov 2010 16:50:07 +0100 Subject: [Beowulf] Intel: 1,000-core processor possible Message-ID: <20101122155007.GI9434@leitl.org> http://www.goodgearguide.com.au/article/368762/intel_1_000-core_processor_possible/ Intel: 1,000-core processor possible A group of Intel researchers has pioneered a messaging system that would allow multiple cores to communicate Joab Jackson (IDG News Service)20 November, 2010 09:04 An experimental Intel chip shows the feasibility of building processors with 1,000 cores, an Intel researcher has asserted. The architecture for the Intel 48-core Single Chip Cloud Computer[1] (SCC) processor is "arbitrarily scalable," said Intel researcher Timothy Mattson, during a talk at the Supercomputer 2010 conference being held this week in New Orleans. "This is an architecture that could, in principle, scale to 1,000 cores," he said. " I can just keep adding, adding, adding cores." Only after 1,000 cores or so, the diameter of the mesh, or the on-chip network connecting the many cores, will grow to such an extent that it would negatively impact performance, Mattson said. Intel remains adamant that the future progress of microprocessors will depend on packing ever more cores onto a chip. As more cores are added, however, Intel designers must confront the problem of scalability. Initial multicore chip architectures depended on a set of protocols that assures that each core has the same view of the system's memory, a technique called cache coherency. As more cores are added to chips, this approach becomes problematic insofar that "the protocol overhead per core grows with the number of cores, leading to a 'coherency wall' beyond which the overhead exceeds the value of adding cores," the paper accompanying Mattson's talk noted. Mattson has argued[2] that a better approach would be to eliminate cache coherency and instead allow cores to pass messages among one another. The recent work of the design team has centered on developing message-passing techniques for the chip that would scale as more cores are added. Designed by Intel's TeraScale Research Program over the past several years, the chip itself is an experimental one and is not on the Intel product road map, Mattson said. A limited number of copies have been distributed to researchers and developers so they can build development tools for the design. The chip, first fabricated with a 45-nanometer process at Intel facilities about a year ago, is actually a six-by-four array of tiles, each tile containing two cores. It has more than 1.3 billion transistors and consumes from 25 to 125 watts. For simplicity's sake, the team used an off-the-shelf 1994-era Pentium processor design for the cores themselves. "Performance on this chip is not interesting," Mattson said. It uses a standard x86 instruction set. The novelty of this processor is in its tiled architecture and the network and address infrastructure. Each core has a "mesh interface component" that packages data into packets and connects to an on-board router. Each tile also has a "message-passing buffer," with 16 kilobytes of random access memory. The team has tried various approaches to streamline the ability of the processor to pass messages among the many cores. By installing the TCP/IP protocol on the data link layer, the team was able to run a separate Linux-based operating system on each core. Mattson noted that while it would be possible to run a 48-node Linux cluster on the chip, it "would be boring." "To make this interesting, I would have to ask, how would the programming models map onto the unique features of this chip," he said. The team also developed a small API (application programming interface) library for message passing among the cores, called RCCE, and which Mattson pronounced as "Rocky." In tests, the team showed that message passing among the cores could be just as speedy using RCCE as with TCP/IP-based Linux cluster. And both approaches bode well for the message-passing approach for inter-core communication. "Our preliminary work has demonstrated that the SCC processor and its native message passing API provide an effective software development platform," the paper concludes. "The expected difficulties due to the lack of asynchronous message passing have so far not materialized." In addition to talking about the chip's message-passing capabilities, Mattson also elaborated on SCC's power-saving capabilities[3]. The frequency of each tile can be varied. Hooks are provided for programmers that would allow their programs to adjust the frequency speed and even the voltage of the cores they are running upon. This feature will, however, create a new challenge for programmers, he warned. "It's a lot harder than you'd think to look at your program and think 'how many volts do I really need?'" he said. Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson[4]. Joab's e-mail address is Joab_Jackson at idg.com[5] References http://www.computerworld.com/s/article/9174981/Intel_to_ship_samples_of_experimental_48_core_chip http://og-hpc.com/Rice2010/Slides/Mattson-OG-HPC-2010-Intel.pdf http://www.pcworld.com/businesscenter/article/183653/intel_48core_singlechip_cloud_computer_improves_power_efficiency.html http://twitter.com/Joab_Jackson mailto:Joab_Jackson at idg.com _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From gus at ldeo.columbia.edu Mon Nov 22 11:42:42 2010 From: gus at ldeo.columbia.edu (Gus Correa) Date: Mon, 22 Nov 2010 11:42:42 -0500 Subject: [Beowulf] (no subject) In-Reply-To: References: Message-ID: <4CEA9D82.3020408@ldeo.columbia.edu> Would PETSc be of use for you? http://www.mcs.anl.gov/petsc/petsc-as/ Gus Correa Ivan Marin wrote: > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I > did some tests in the past for simple stuff, and worked, but very > simple linear system solvers. > > Ivan Marin > > Civil Engineering Dept > University of Minnesota > 500 Pillsbury Dr. S.E. > Minneapolis, MN 55455 > > Laborat?rio de Hidr?ulica Computacional - LHC > Departamento de Hidr?ulica e Saneamento - SHS > Escola de Engenharia de S?o Carlos - EESC > Universidade de S?o Paulo - USP > > http://albatroz.shs.eesc.usp.br > +55 16 3373 8270 > > > > 2010/11/17 Peter St. John : >> I'm going to dabble a bit with writing some C/C++ with a linear algebra >> library. My fortran is rusty (although I did some work translating fortran >> to C in this millennium :-). >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice have >> any real bearing on compiling against MPI later? >> Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) I'd >> be glad to hear it. >> >> I just installed (k)ubuntu, which seems fine for my purposes so far. I do >> miss fvwm, and the default "single click to invoke an application" is >> totally nuts but I fixed it. >> Thanks, >> Peter >> >> _______________________________________________ >> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing >> To change your subscription (digest mode or unsubscribe) visit >> http://www.beowulf.org/mailman/listinfo/beowulf >> >> > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Mon Nov 22 12:02:20 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Mon, 22 Nov 2010 12:02:20 -0500 Subject: [Beowulf] ("LAPACK/C++ wrappers) Message-ID: Gus, In my case, I"m not concerned with PDEs but the "linear system solvers" themselves,which PETSci gets from other packages? So for example, looking over their table at: http://www.mcs.anl.gov/petsc/petsc-as/documentation/linearsolvertable.html I'd want the "dense" case and therefore PLAPACK, so now I know to google PLAPACK. I just want to solve a system of simultaneous linear equations within the vernacular of C/C++. Thanks, Peter On Mon, Nov 22, 2010 at 11:42 AM, Gus Correa wrote: > Would PETSc be of use for you? > > http://www.mcs.anl.gov/petsc/petsc-as/ > > Gus Correa > > Ivan Marin wrote: > > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I > > did some tests in the past for simple stuff, and worked, but very > > simple linear system solvers. > > > > Ivan Marin > > > > Civil Engineering Dept > > University of Minnesota > > 500 Pillsbury Dr. S.E. > > Minneapolis, MN 55455 > > > > Laborat?rio de Hidr?ulica Computacional - LHC > > Departamento de Hidr?ulica e Saneamento - SHS > > Escola de Engenharia de S?o Carlos - EESC > > Universidade de S?o Paulo - USP > > > > http://albatroz.shs.eesc.usp.br > > +55 16 3373 8270 > > > > > > > > 2010/11/17 Peter St. John : > >> I'm going to dabble a bit with writing some C/C++ with a linear algebra > >> library. My fortran is rusty (although I did some work translating > fortran > >> to C in this millennium :-). > >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice > have > >> any real bearing on compiling against MPI later? > >> Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) > I'd > >> be glad to hear it. > >> > >> I just installed (k)ubuntu, which seems fine for my purposes so far. I > do > >> miss fvwm, and the default "single click to invoke an application" is > >> totally nuts but I fixed it. > >> Thanks, > >> Peter > >> > >> _______________________________________________ > >> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin > Computing > >> To change your subscription (digest mode or unsubscribe) visit > >> http://www.beowulf.org/mailman/listinfo/beowulf > >> > >> > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From gus at ldeo.columbia.edu Mon Nov 22 12:38:19 2010 From: gus at ldeo.columbia.edu (Gus Correa) Date: Mon, 22 Nov 2010 12:38:19 -0500 Subject: [Beowulf] ("LAPACK/C++ wrappers) In-Reply-To: References: Message-ID: <4CEAAA8B.8060702@ldeo.columbia.edu> Peter St. John wrote: > Gus, > In my case, I"m not concerned with PDEs but the "linear system solvers" > themselves,which PETSci gets from other packages? Hi Peter Yes, you can compile PETSC with support of a variety of linear algebra packages. See their installation instructions and documentation. I think there is support for PLAPACK, please check their install and docs pages: http://www.mcs.anl.gov/petsc/petsc-2/documentation/installation.html http://www.mcs.anl.gov/petsc/petsc-2/documentation/index.html > So for example, > looking over their table at: > http://www.mcs.anl.gov/petsc/petsc-as/documentation/linearsolvertable.html > I'd want the "dense" case and therefore PLAPACK, so now I know to > google PLAPACK. > I just want to solve a system of simultaneous linear equations within > the vernacular of C/C++. AFAIK, PETSc has C, Fortran, and C++ interfaces. If I remember right, you can choose between C++ or C when you compile it. You can choose between real (for most linear algebra ) or complex (a must for FFTs, and FFTW support) scalars. You can also choose which linear algebra packages you will use (which you can compile separately, or let PETSc handle - it handles many of them), which BLAS/LAPACK to use (say, Goto BLAS, Intel MKL, AMD ACML, or let PETSc install its own version). I am not a PETSc user, but some people here used PETSc to solve inverse and forward problems, problems with sparse and dense matrices, etc. As it often happens when science (not computer science) is the main goal, once the "it works" stage is reached, the code starts to be used to produce papers, more and more features are added, and it is never again improved for efficiency, never goes beyond the prototype phase in this regard. In terms of efficiency, the results didn't shine, in terms of science/results output they were very good. However, it may well be that the way the problem (not PETSc) was programmed led to the inefficient code. PETSc experts in the list: For the benefit of Peter, Ivan, and the list, would you kindly jump in and say something about PETSc efficiency and other features? Txs. Maybe PETSc is an overkill for you, but if the problem is that simple, why not use Fortran and regular LAPACK/PLAPACK? My two cents. Gus Correa > Thanks, > Peter > > On Mon, Nov 22, 2010 at 11:42 AM, Gus Correa > wrote: > > Would PETSc be of use for you? > > http://www.mcs.anl.gov/petsc/petsc-as/ > > Gus Correa > > Ivan Marin wrote: > > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I > > did some tests in the past for simple stuff, and worked, but very > > simple linear system solvers. > > > > Ivan Marin > > > > Civil Engineering Dept > > University of Minnesota > > 500 Pillsbury Dr. S.E. > > Minneapolis, MN 55455 > > > > Laborat?rio de Hidr?ulica Computacional - LHC > > Departamento de Hidr?ulica e Saneamento - SHS > > Escola de Engenharia de S?o Carlos - EESC > > Universidade de S?o Paulo - USP > > > > http://albatroz.shs.eesc.usp.br > > +55 16 3373 8270 > > > > > > > > 2010/11/17 Peter St. John >: > >> I'm going to dabble a bit with writing some C/C++ with a linear > algebra > >> library. My fortran is rusty (although I did some work > translating fortran > >> to C in this millennium :-). > >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this > choice have > >> any real bearing on compiling against MPI later? > >> Just if anyone is happy with any combination (LAPACK++ and > OpenMPI, say) I'd > >> be glad to hear it. > >> > >> I just installed (k)ubuntu, which seems fine for my purposes so > far. I do > >> miss fvwm, and the default "single click to invoke an > application" is > >> totally nuts but I fixed it. > >> Thanks, > >> Peter > >> > >> _______________________________________________ > >> Beowulf mailing list, Beowulf at beowulf.org > sponsored by Penguin Computing > >> To change your subscription (digest mode or unsubscribe) visit > >> http://www.beowulf.org/mailman/listinfo/beowulf > >> > >> > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org > sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org > sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Mon Nov 22 13:35:10 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Mon, 22 Nov 2010 13:35:10 -0500 Subject: [Beowulf] ("LAPACK/C++ wrappers) In-Reply-To: <4CEAAA8B.8060702@ldeo.columbia.edu> References: <4CEAAA8B.8060702@ldeo.columbia.edu> Message-ID: Regarding "...Maybe PETSc is an overkill for you, but if the problem is that simple, why not use Fortran and regular LAPACK/PLAPACK?,,," I haven't written fortran since IV and 77 in the early 80's. The matrix multiplication part is simple but the ambient application is squirrelly and I"m more facile in C or C++. So I just want basic linear algebra library I can link my C to, but maybe all y'all use Matlab for simple things. Probably I should just do it with, say, Armadillo, and learn something without spilling too much blood from fingertips. Peter On Mon, Nov 22, 2010 at 12:38 PM, Gus Correa wrote: > Peter St. John wrote: > > Gus, > > In my case, I"m not concerned with PDEs but the "linear system solvers" > > themselves,which PETSci gets from other packages? > Hi Peter > > Yes, you can compile PETSC with support of a variety of linear algebra > packages. See their installation instructions and documentation. > I think there is support for PLAPACK, please check their > install and docs pages: > > http://www.mcs.anl.gov/petsc/petsc-2/documentation/installation.html > http://www.mcs.anl.gov/petsc/petsc-2/documentation/index.html > > > So for example, > > looking over their table at: > > > http://www.mcs.anl.gov/petsc/petsc-as/documentation/linearsolvertable.html > > I'd want the "dense" case and therefore PLAPACK, so now I know to > > google PLAPACK. > > I just want to solve a system of simultaneous linear equations within > > the vernacular of C/C++. > > AFAIK, PETSc has C, Fortran, and C++ interfaces. > If I remember right, you can choose between C++ or C when > you compile it. > You can choose between real (for most linear algebra ) > or complex (a must for FFTs, and FFTW support) scalars. > You can also choose which > linear algebra packages you will use (which you can compile separately, > or let PETSc handle - it handles many of them), > which BLAS/LAPACK to use (say, Goto BLAS, > Intel MKL, AMD ACML, or let PETSc install its own version). > > I am not a PETSc user, but some people here used PETSc to solve inverse > and forward problems, problems with sparse and dense matrices, etc. > As it often happens when science (not computer science) > is the main goal, once the "it works" stage is reached, > the code starts to be used to produce papers, > more and more features are added, > and it is never again improved for efficiency, > never goes beyond the prototype phase in this regard. > > In terms of efficiency, the results didn't shine, > in terms of science/results output they were very good. > However, it may well be that the way the problem > (not PETSc) was programmed led to the inefficient code. > > PETSc experts in the list: > For the benefit of Peter, Ivan, and the list, > would you kindly jump in and say something about PETSc > efficiency and other features? > Txs. > > Maybe PETSc is an overkill for you, but if the problem is > that simple, why not use Fortran and regular LAPACK/PLAPACK? > > My two cents. > Gus Correa > > > Thanks, > > Peter > > > > On Mon, Nov 22, 2010 at 11:42 AM, Gus Correa > > wrote: > > > > Would PETSc be of use for you? > > > > http://www.mcs.anl.gov/petsc/petsc-as/ > > > > Gus Correa > > > > Ivan Marin wrote: > > > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. > I > > > did some tests in the past for simple stuff, and worked, but very > > > simple linear system solvers. > > > > > > Ivan Marin > > > > > > Civil Engineering Dept > > > University of Minnesota > > > 500 Pillsbury Dr. S.E. > > > Minneapolis, MN 55455 > > > > > > Laborat?rio de Hidr?ulica Computacional - LHC > > > Departamento de Hidr?ulica e Saneamento - SHS > > > Escola de Engenharia de S?o Carlos - EESC > > > Universidade de S?o Paulo - USP > > > > > > http://albatroz.shs.eesc.usp.br > > > +55 16 3373 8270 > > > > > > > > > > > > 2010/11/17 Peter St. John > >: > > >> I'm going to dabble a bit with writing some C/C++ with a linear > > algebra > > >> library. My fortran is rusty (although I did some work > > translating fortran > > >> to C in this millennium :-). > > >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this > > choice have > > >> any real bearing on compiling against MPI later? > > >> Just if anyone is happy with any combination (LAPACK++ and > > OpenMPI, say) I'd > > >> be glad to hear it. > > >> > > >> I just installed (k)ubuntu, which seems fine for my purposes so > > far. I do > > >> miss fvwm, and the default "single click to invoke an > > application" is > > >> totally nuts but I fixed it. > > >> Thanks, > > >> Peter > > >> > > >> _______________________________________________ > > >> Beowulf mailing list, Beowulf at beowulf.org > > sponsored by Penguin Computing > > >> To change your subscription (digest mode or unsubscribe) visit > > >> http://www.beowulf.org/mailman/listinfo/beowulf > > >> > > >> > > > _______________________________________________ > > > Beowulf mailing list, Beowulf at beowulf.org > > sponsored by Penguin Computing > > > To change your subscription (digest mode or unsubscribe) visit > > http://www.beowulf.org/mailman/listinfo/beowulf > > > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org > > sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > > http://www.beowulf.org/mailman/listinfo/beowulf > > > > > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From ispmarin at gmail.com Tue Nov 23 11:48:37 2010 From: ispmarin at gmail.com (Ivan Marin) Date: Tue, 23 Nov 2010 10:48:37 -0600 Subject: [Beowulf] C/C++ Numerical Integral solvers In-Reply-To: <4CEAAA8B.8060702@ldeo.columbia.edu> References: <4CEAAA8B.8060702@ldeo.columbia.edu> Message-ID: <4CEBF065.4070809@gmail.com> Hi folks, I'm following closely the thread about the numerical solvers in C++, and would like to ask: is anyone using or has experience with numerical integral solvers, hopefully parallel, in C++? I'm solving singular Cauchy type integrals, complex. I've tried the GSL and they are very precise, but _very_ slow for my type of integrands, and not intrinsically threaded/parallel. -- Ivan Marin Civil Engineering Dept University of Minnesota 500 Pillsbury Dr. S.E. Minneapolis, MN 55455 Laborat?rio de Hidr?ulica Computacional - LHC Departamento de Hidr?ulica e Saneamento - SHS Escola de Engenharia de S?o Carlos - EESC Universidade de S?o Paulo - USP http://albatroz.shs.eesc.usp.br +55 16 3373 8270 _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Mon Nov 1 01:19:58 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Mon, 1 Nov 2010 01:19:58 -0400 Subject: [Beowulf] "Go" programming language Message-ID: Has anyone tried the "Go" programming language on a beowulf? The language's homepage says, " Its concurrency mechanisms make it easy to write programs that get the most out of multicore and networked machines..." (from http://golang.org/) The wiki is http://en.wikipedia.org/wiki/Go_%28programming_language%29 I'm sure I'll use MPI but Google hired some pretty cool language designers. Peter P.S. described somewhere as "merging C++ with Python" which maybe explains an odd white-space rule (open curly bracket can't begin a line because it would confuse automated semicolon line endings), Yuk. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From deadline at eadline.org Mon Nov 1 09:36:48 2010 From: deadline at eadline.org (Douglas Eadline) Date: Mon, 1 Nov 2010 09:36:48 -0400 (EDT) Subject: [Beowulf] SC10 Beowulf Bash Short Interview In-Reply-To: <43163.192.168.93.213.1288401573.squirrel@mail.eadline.org> References: <68A57CCFD4005646957BD2D18E60667B121AD077@milexchmb1.mil.tagmclarengroup.com> <43163.192.168.93.213.1288401573.squirrel@mail.eadline.org> Message-ID: <34986.192.168.93.213.1288618608.squirrel@mail.eadline.org> For those that may be wondering, there is some background on the Beowulf Bash at: http://insidehpc.com/2010/11/01/monday-beowulf-bash-is-the-big-easy-at-sc10/ -- Doug > > It is that time of the year. If you are attending SC10, > here is what you have been waiting for: > > The Big Wheels Keep On Turning Beowulf Bash > > http://www.xandmarketing.com/beobash10/ > > -- > Doug > > -- > This message has been scanned for viruses and > dangerous content by MailScanner, and is > believed to be clean. > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- Doug -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From Bill.Rankin at sas.com Mon Nov 1 11:21:36 2010 From: Bill.Rankin at sas.com (Bill Rankin) Date: Mon, 1 Nov 2010 15:21:36 +0000 Subject: [Beowulf] RE: Storage - the end of RAID? In-Reply-To: <20101029191505.GA29737@bx9.net> References: <68A57CCFD4005646957BD2D18E60667B121AD077@milexchmb1.mil.tagmclarengroup.com> <4CCB168B.6000002@runnersroll.com> <20101029191505.GA29737@bx9.net> Message-ID: <76097BB0C025054786EFAB631C4A2E3C0949E2E8@MERCMBX03R.na.SAS.com> > Um, it's not really RAID 1 when the drives are in different servers. > Although there's not much point in arguing about that. > > -- greg My knee-jerk reaction to Greg's statement was going to be something snotty and along the lines of "what part of 'R' don't you understand?" ;-) But upon further pondering and also going back and re-reading the original article a little more slowly, I think that there is a point in there. Now, one thing I do have an issue with is the article's title claim that RAID is somehow dead. This is clarified in the very first sentence where he identifies that this refers to "costly RAID controllers". Terribly misleading but as someone mentioned earlier it does grab your attention. But in a sense this is old news to us. What the article essentially addresses is that with the huge increase in the I/O capability of other pieces of the system (CPUs, busses, etc) the model of having all your data accessed by going out on the wire and pulling it in from some remote (possibly multiple) high end storage servers cannot survive. At least not for "active" data. Well, duh. Personally I've been using clusters to crush some decent high-end storage arrays since around 2003. Back then the general rule was that we had local disk on the nodes and you would do any significant I/O to those disks. We would stage static copies of input data (eg. Genomics databases) to those disks also just to avoid going out on the wire. So we threw out RAID years ago for our active data. Where we did keep it was for our less active data - home directories, executables, etc - where we did have a need for a definitive, coherent storage image and availability was more important that absolute performance. But this too is starting to change as the node count rises. I think that the current thinking is that while disks have gotten very large, their I/O performance has not kept pace. With the cost of solid state storage coming down in price now, it makes a lot of sense to start replacing disks where we have single point bottlenecks in our I/O chain. So I look at the whole discussion as the realization that finally the rest of the world is catching up to us. :-) -bill _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rtomek at ceti.pl Mon Nov 1 12:24:55 2010 From: rtomek at ceti.pl (Tomasz Rola) Date: Mon, 1 Nov 2010 17:24:55 +0100 (CET) Subject: [Beowulf] "Go" programming language In-Reply-To: References: Message-ID: On Mon, 1 Nov 2010, Peter St. John wrote: > Has anyone tried the "Go" programming language on a beowulf? Probably not, I wildly guess. First of all, the language is too new (just one year old). Besides, it looks like they wanted to have a language that didn't look like C, even thou it looked like it. And they succeded... > The language's homepage says, > " Its concurrency mechanisms make it easy to write programs that get the > most out of multicore and networked machines..." Well, myself, maybe I would stand on my hands, clap my feet and yell "ooom! oom!" if they did this to Haskell (maybe there is some library for distributing Haskell, I'm not sure, had no time/need to check). Personally, I don't think I would use Go. At least in the foreseeable future. It doesn't seem it gives anything that Python and C couldn't give, and they are better established as languages. At best, it may be interesting syntactic experiment. If few years from now there is still some rumour about it, then maybe... > (from http://golang.org/) > > The wiki is http://en.wikipedia.org/wiki/Go_%28programming_language%29 > > I'm sure I'll use MPI but Google hired some pretty cool language designers. Yes. There is a proverb in Polish, that roughly translates as "when you hire six chefs in a kitchen, you have nothing to eat". > Peter > > P.S. described somewhere as "merging C++ with Python" which maybe > explains an odd white-space rule (open curly bracket can't begin a line > because it would confuse automated semicolon line endings), Yuk. Genetically-wise, some hybrids are great and move things forward. But some other are unable to live without life support. Regards, Tomasz Rola -- ** A C programmer asked whether computer had Buddha's nature. ** ** As the answer, master did "rm -rif" on the programmer's home ** ** directory. And then the C programmer became enlightened... ** ** ** ** Tomasz Rola mailto:tomasz_rola at bigfoot.com ** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Mon Nov 1 12:37:36 2010 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Mon, 1 Nov 2010 09:37:36 -0700 Subject: [Beowulf] RE: Storage - the end of RAID? In-Reply-To: <76097BB0C025054786EFAB631C4A2E3C0949E2E8@MERCMBX03R.na.SAS.com> References: <68A57CCFD4005646957BD2D18E60667B121AD077@milexchmb1.mil.tagmclarengroup.com> <4CCB168B.6000002@runnersroll.com> <20101029191505.GA29737@bx9.net> <76097BB0C025054786EFAB631C4A2E3C0949E2E8@MERCMBX03R.na.SAS.com> Message-ID: > I think that the current thinking is that while disks have gotten very large, their I/O performance > has not kept pace. With the cost of solid state storage coming down in price now, it makes a lot of > sense to start replacing disks where we have single point bottlenecks in our I/O chain. Your observation isn't surprising.. over time, density increases (on the disk, on silicon, heck, on punched cards), but the data still has to flow serially through some medium (even if you have a parallel databus, it's not megabits wide), and there, you're limited by EM propagation and all its ills. Speed over a wire, unfortunately, doesn't increase exponentially like density does. It doesn't even get a square law for feature spacing vs areal density. The other issue that bites you pretty hard is power consumption. Fast data = more transitions = more units of charge moving from one place to another = more IR losses. And that's without looking at the more complex transmitting/receiving hardware needed as you move from pushbuttons and relays to things that have to worry about impedance discontinuities and adaptive equalization. In many ways, the whole idea of distributed computing is equally applicable to distributed storage, problems and all, just a matter of the scale whether it's registers in the CPU, cache, some level of RAM, or bulk storage. > > So I look at the whole discussion as the realization that finally the rest of the world is catching up > to us. > In many ways, Beowulfery has helped here.. especially in its early incarnations, the "between node" pipes were pretty slow compared to previous supercomputer designs, so people spent a lot of time figuring out how to structure algorithms so they had good locality of reference and were decoupled at fine time scales. It's sort of the inverse of the classic array processor, systolic array, or even SIMD machines. I *like* having architectures generically based on message passing rather than shared memory. Programming is harder at first, because you need to explicitly recognize the non-deterministic behavior of the messages, but I think it makes the result design cleaner from an architectural standpoint. It really gets rid of the "global shared variable" thing that is a bane of multithreaded programming. (Of course, if you're coming from a tightly coupled environment with fast semaphores, you find it a pain.. ) _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From nixon at nsc.liu.se Tue Nov 2 08:39:27 2010 From: nixon at nsc.liu.se (Leif Nixon) Date: Tue, 02 Nov 2010 13:39:27 +0100 Subject: [Beowulf] Anybody using Redhat HPC Solution in their Beowulf In-Reply-To: <4CC6FD28.1050303@runnersroll.com> (Ellis H. Wilson, III's message of "Tue, 26 Oct 2010 12:09:12 -0400") References: <4CC11EA8.8030602@gmail.com> <68A57CCFD4005646957BD2D18E60667B12154605@milexchmb1.mil.tagmclarengroup.com> <4CC60BFA.4080502@runnersroll.com> <68A57CCFD4005646957BD2D18E60667B12154E23@milexchmb1.mil.tagmclarengroup.com> <4CC6FD28.1050303@runnersroll.com> Message-ID: "Ellis H. Wilson III" writes: > Optimally IMHO, in university setups physical scientists create the > need for HPC. These types shouldn't (as Kilian mentions) need to > inherit all of the responsibilities and overheads of cluster > management to use one (or pay cluster vendors annually for support). > They should simply walk over to the CS department Agh, no, not the CS guys. They should come to their friendly local HPC centre, who will help them spec, acquire and operate a custom cluster (or give them a time allocation on a shared cluster). I'm being purely objective here, of course. -- / Swedish National Infrastructure for Computing Leif Nixon - Security officer < National Supercomputer Centre \ Nordic Data Grid Facility _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From tomislav.maric at gmx.com Tue Nov 2 17:45:51 2010 From: tomislav.maric at gmx.com (tomislav_maric@gmx.com) Date: Tue, 02 Nov 2010 22:45:51 +0100 Subject: [Beowulf] cluster profiling Message-ID: <20101102215301.225020@gmx.com> Hi everyone, I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". Best regards, Tomislav Maric, (MSc Mechanical Engineering, just to clarify my ignorance regarding HPC) _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From hearnsj at googlemail.com Tue Nov 2 19:21:13 2010 From: hearnsj at googlemail.com (John Hearns) Date: Tue, 2 Nov 2010 23:21:13 +0000 Subject: [Beowulf] cluster profiling In-Reply-To: <20101102215301.225020@gmx.com> References: <20101102215301.225020@gmx.com> Message-ID: On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > This sounds very cool. To be honest, most people use Excel spreadsheets to plot this sort of thing. If you can produce an automated framework to do this it would be very interesting. I have to slightly question your choice of Ganglia - have you thought of using sysstat to capture the system's load or memory figures? _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Tue Nov 2 20:49:34 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Tue, 2 Nov 2010 20:49:34 -0400 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: deBeers laser-engraves serial numbers onto their (natural) diamonds (to counter the increasing gem quality of artificial diamonds made by, say, chemical vapor deposition). So how about laser engraving data onto cheap chemical vapor deposition thin diamond slices? (One of the ideas had been to make microelectronics substrates from diamond this way, since diamond conducts heat better than silicon). Peter On Fri, Oct 29, 2010 at 1:36 PM, Robert G. Brown wrote: > On Fri, 29 Oct 2010, Lux, Jim (337C) wrote: > > Or, how about something like the UNICON aka "terabit memory" (TBM) from >> Illiac IV days. It's a stable polyester base with a thin film of rhodium >> that was ablated by a laser making 3 micron holes to write the bits. >> $3.5M >> to store a terabit in 1975. >> > > Burned RO laser disks should in principle be as stable, if the medium > used is thick enough. The problem is that CDs tend to be mass produced > with very thin media, cheap plastic, and are even susceptible to > corrosion through the plastic over time. If one made a CD with tempered > glass and a moderately thick slice of e.g. stainless steel or > platinum... > > But then your problem is the reader. CD readers give way to DVD and are > still backwards compatible, sort of. But what about the 2020 > equivalent? Will there even be one? Nobody will buy actual CDs any > more. Nobody will buy movies on DVDs any more (seriously, I doubt that > they will). Will there BE a laser drive that is backwards compatible to > CD, or will it go the way of reel to reel tapes, 8 track tapes, cassette > tapes, QIC tapes, floppy drives of all flavors (including high capacity > drives like the ones I have carefully saved at home in case I ever need > one), magnetic core memories, large mountable disk packs, exabyte tape > drives, DA tapes, and so on? I rather think it will be gone. It isn't > even clear if hard disk drives will still be available (not that any > computer around would be able to interface with the 5 or 10 MB drives of > my youth anyway). > > This is the problem with electronics. You have to have BOTH long time > scale stability AND an interface for the ages. And the latter is highly > incompatible with e.g. Moore's Law -- not even the humble serial port > has made it through thirty years unscathed. Is the Universal Serial Bus > really Universal? I doubt it. And yet, that is likely to be the only > interface available AT ALL (except for perhaps some sort of wireless > network that isn't even VISIBLE to old peripherals) on the vast bulk of > the machines sold in a mere five years. > > A frightening trend in computing these days is that we may be peaking in > the era where one's computer (properly equipped with a sensible > operating system) is symmetrically capable of functioning as a client > and a server. Desktop computers were clients, servers, or both as one > wished, from the days of Sun workstations through to the present, with > any sort of Unixoid operating system and adequate resources. From the > mid 90's on, with Linux, pure commodity systems were both at the whim of > the system owner -- anybody could add more memory, more disks, a backup > device, and the same chassis was whatever you needed it to be. > > Now, however, this general purpose desktop is all but dead, supplanted > by laptops that are just as powerful, but that lack the expandability > and repurposeability. And laptops are themselves an endangered species > all of a sudden -- in five years a "laptop" could very well be a single > "pad" (touchscreen) of whatever size with or without an external > keyboard, all wireless, smooth as a baby's bottom as far as actual plugs > are concerned (or maybe, just maybe, with a single USB charger/data port > or a couple of slots for SD-of-the-day or USB peripherals). Actual data > storage may well migrate into servers that are completely different > beasts, far away, accessible only over a wireless network, and > controlled by others. > > An enormous step backwards, in other words. A risk to our political > freedom. And yet so seductive, so economical, so convenient, that we > may willingly dance down a primrose path to an information catastrophe > that is more or less impossible still with the vast decentralization of > stored knowledge. > > rgb > > > Robert G. Brown http://www.phy.duke.edu/~rgb/ > Duke University Dept. of Physics, Box 90305 > Durham, N.C. 27708-0305 > Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu > > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From samuel at unimelb.edu.au Tue Nov 2 20:49:59 2010 From: samuel at unimelb.edu.au (Christopher Samuel) Date: Wed, 03 Nov 2010 11:49:59 +1100 Subject: [Beowulf] cluster profiling In-Reply-To: <20101102215301.225020@gmx.com> References: <20101102215301.225020@gmx.com> Message-ID: <4CD0B1B7.9020705@unimelb.edu.au> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 03/11/10 08:45, tomislav_maric at gmx.com wrote: > that will use the Ganglia-python interface and try to > give me an insight into the way machine is burdened > during runs Depending on how old your kernel is the "perf" utility (found in the tools/perf directory in your kernel sources, or packaged in Ubuntu as part of the linux-tools package or linux-tools-2.6 in Debian Squeeze) may well give you some interesting stats. As a here is an overview of stats a "find -ls" over the current kernel git tree: $ perf stat find . -ls > /dev/null Performance counter stats for 'find . -ls': 372.415331 task-clock-msecs # 0.923 CPUs 158 context-switches # 0.000 M/sec 2 CPU-migrations # 0.000 M/sec 395 page-faults # 0.001 M/sec 648855865 cycles # 1742.291 M/sec 698863597 instructions # 1.077 IPC 14321645 cache-references # 38.456 M/sec 379109 cache-misses # 1.018 M/sec 0.403454703 seconds time elapsed You can use the "perf list" command to get a list of all the kernel tracepoints you can monitor and then you can select them individually with the "stat" command. Here is perf monitoring CPU migrations, L1 dcache misses and the kernel scheduler stats of that well known HPC program "top". ;-) perf stat -e migrations -e L1-dcache-load-misses -e sched:* top [...] Performance counter stats for 'top': 0 CPU-migrations # 0.000 M/sec 1038307 L1-dcache-load-misses # 0.000 M/sec 0 sched:sched_kthread_stop # 0.000 M/sec 0 sched:sched_kthread_stop_ret # 0.000 M/sec 0 sched:sched_wait_task # 0.000 M/sec 98 sched:sched_wakeup # 0.000 M/sec 0 sched:sched_wakeup_new # 0.000 M/sec 61 sched:sched_switch # 0.000 M/sec 0 sched:sched_migrate_task # 0.000 M/sec 0 sched:sched_process_free # 0.000 M/sec 1 sched:sched_process_exit # 0.000 M/sec 0 sched:sched_process_wait # 0.000 M/sec 0 sched:sched_process_fork # 0.000 M/sec 15 sched:sched_signal_send # 0.000 M/sec 49 sched:sched_stat_wait # 0.000 M/sec 174 sched:sched_stat_runtime # 0.000 M/sec 67 sched:sched_stat_sleep # 0.000 M/sec 0 sched:sched_stat_iowait # 0.000 M/sec 29.452075124 seconds time elapsed With root access you can even do "perf top" to see what's going on under the hood. You can also use "perf record -g $COMMAND" to record the profiling information for $COMMAND to perf.data along with call graph information so you can display a detailed tree view of what was going on via the "perf report" command. Quite a neat little tool I've got to say! cheers, Chris - -- Christopher Samuel - Senior Systems Administrator VLSCI - Victorian Life Sciences Computational Initiative Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 http://www.vlsci.unimelb.edu.au/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkzQsbcACgkQO2KABBYQAh+6JACaAx7p0zARcGGO4busVv7AbqHL tCcAnA4Z6HOs1LTbucprnyBJFxF6glo+ =D2wX -----END PGP SIGNATURE----- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Tue Nov 2 21:19:22 2010 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Tue, 2 Nov 2010 18:19:22 -0700 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: Diamonds may be almost forever, But I would think that fused silica would work almost as well, and is substantially less expensive. If you want something exotic, how about ion implantation of Cr+ or Ti+ ions into alumina Jim Lux +1(818)354-2075 From: Peter St. John [mailto:peter.st.john at gmail.com] Sent: Tuesday, November 02, 2010 5:50 PM To: Robert G. Brown Cc: Lux, Jim (337C); beowulf at beowulf.org Subject: Re: [Beowulf] Re: Interesting deBeers laser-engraves serial numbers onto their (natural) diamonds (to counter the increasing gem quality of artificial diamonds made by, say, chemical vapor deposition). So how about laser engraving data onto cheap chemical vapor deposition thin diamond slices? (One of the ideas had been to make microelectronics substrates from diamond this way, since diamond conducts heat better than silicon). Peter -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From samuel at unimelb.edu.au Tue Nov 2 22:15:11 2010 From: samuel at unimelb.edu.au (Christopher Samuel) Date: Wed, 03 Nov 2010 13:15:11 +1100 Subject: [Beowulf] "Go" programming language In-Reply-To: References: Message-ID: <4CD0C5AF.5000806@unimelb.edu.au> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 02/11/10 03:24, Tomasz Rola wrote: > Well, myself, maybe I would stand on my hands, clap my feet and yell > "ooom! oom!" if they did this to Haskell (maybe there is some library for > distributing Haskell, I'm not sure, had no time/need to check). One of our guys (Bernie Pope) who is a FP geek has been working on MPI bindings for Haskell in his free time (the older ones had been abandoned some time ago). It was presented at the AusHac Australian Haskell Hackathon this July. Presentation here: http://www.berniepope.id.au/docs/mpi_bindings.pdf Code here: http://github.com/bjpop/haskell-mpi cheers! Chris - -- Christopher Samuel - Senior Systems Administrator VLSCI - Victorian Life Sciences Computational Initiative Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 http://www.vlsci.unimelb.edu.au/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkzQxa4ACgkQO2KABBYQAh8LowCdFbSahWYbdMJR0/Ge9QLKaDxD f8MAn2YdYXuCG/zjnh5pDQBrtdo9BWjg =VXop -----END PGP SIGNATURE----- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rtomek at ceti.pl Tue Nov 2 23:06:30 2010 From: rtomek at ceti.pl (Tomasz Rola) Date: Wed, 3 Nov 2010 04:06:30 +0100 (CET) Subject: [Beowulf] "Go" programming language In-Reply-To: <4CD0C5AF.5000806@unimelb.edu.au> References: <4CD0C5AF.5000806@unimelb.edu.au> Message-ID: On Wed, 3 Nov 2010, Christopher Samuel wrote: > One of our guys (Bernie Pope) who is a FP geek has been working > on MPI bindings for Haskell in his free time (the older ones had > been abandoned some time ago). It was presented at the AusHac > Australian Haskell Hackathon this July. Wow, interesting. Thanks! Regards, Tomasz Rola -- ** A C programmer asked whether computer had Buddha's nature. ** ** As the answer, master did "rm -rif" on the programmer's home ** ** directory. And then the C programmer became enlightened... ** ** ** ** Tomasz Rola mailto:tomasz_rola at bigfoot.com ** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From ntmoore at gmail.com Wed Nov 3 01:16:30 2010 From: ntmoore at gmail.com (Nathan Moore) Date: Wed, 3 Nov 2010 00:16:30 -0500 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: Diamonds will burn, just like coal. Also, at surface of the earth pressure, diamond is (by free-energy calc) the less stable state of carbon, and over (geologic) time diamond will trend to graphite/coal. Schroeder's "Thermal Physics" text talks about this for a few pages. A friend of mine has a wedding band with a tiny "imperfection" in the stone - a streak of graphite that you can see with a hand-lens. It makes the stone less valuable, but frankly its more interesting than your standard hunk of glitter. On Tue, Nov 2, 2010 at 8:19 PM, Lux, Jim (337C) wrote: > Diamonds may be almost forever, > > But I would think that fused silica would work almost as well, and is > substantially less expensive. > > If you want something exotic, how about ion implantation of Cr+ or Ti+ ions > into alumina > > > > Jim Lux > +1(818)354-2075 > > *From:* Peter St. John [mailto:peter.st.john at gmail.com] > *Sent:* Tuesday, November 02, 2010 5:50 PM > *To:* Robert G. Brown > *Cc:* Lux, Jim (337C); beowulf at beowulf.org > > *Subject:* Re: [Beowulf] Re: Interesting > > > > deBeers laser-engraves serial numbers onto their (natural) diamonds (to > counter the increasing gem quality of artificial diamonds made by, say, > chemical vapor deposition). So how about laser engraving data onto cheap > chemical vapor deposition thin diamond slices? (One of the ideas had been to > make microelectronics substrates from diamond this way, since diamond > conducts heat better than silicon). > Peter > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > -- - - - - - - - - - - - - - - - - - - - - - Nathan Moore Associate Professor, Physics Winona State University - - - - - - - - - - - - - - - - - - - - - -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From hearnsj at googlemail.com Wed Nov 3 04:38:10 2010 From: hearnsj at googlemail.com (John Hearns) Date: Wed, 3 Nov 2010 08:38:10 +0000 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: On 29 October 2010 18:36, Robert G. Brown wrote: > ?Will there BE a laser drive that is backwards compatible to > CD, or will it go the way of reel to reel tapes, 8 track tapes, cassette > tapes, QIC tapes, floppy drives of all flavors (including high capacity > drives like the ones I have carefully saved at home in case I ever need > one), magnetic core memories, large mountable disk packs, exabyte tape > drives, DA tapes, and so on? ?I rather think it will be gone. ?It isn't > even clear if hard disk drives will still be available (not that any > computer around would be able to interface with the 5 or 10 MB drives of > my youth anyway). Have a look at the other thread I started, regarding new approaches to data storage. > > Now, however, this general purpose desktop is all but dead, supplanted... ..... > ?Actual data > storage may well migrate into servers that are completely different > beasts, far away, accessible only over a wireless network, and > controlled by others. > > An enormous step backwards, in other words. ?A risk to our political > freedom. ?And yet so seductive, so economical, so convenient, that we > may willingly dance down a primrose path to an information catastrophe > that is more or less impossible still with the vast decentralization of > stored knowledge. I'm not so sure it IS a step backwards. What I think we should be doing is working towards a media-agnostic form of storing data. A recognition that scientific data (and other forms, like movies and music etc.) will carry with them metadata and that the data will migrate through many types of physical media in its lifetime, and will from the outset have multiple copies made. I guess the HPC Grid computing types are doing this already, what I'm rather thinking about is a universal standard for this, and a way of carrying the metadata with the actual data in a way it cannot be lost. Its also funny that I use the term "lifetime" - I guess in the past we all have assumed digital data will have an infinite lifetime, as as discussed above it has come to pass that the decay of media, or reading apparatus being unavailable has made data have a finite lifetime. The real point I am making here is that with cloud type data storage over IP connections even in HPC we will be seeing data accessed not on SCSI volumes (be that direct SCSI, fibrechannel, iSCSI, RAID etc) but from an HTTP accessed object store. You might then say that "Hey - performance matters and that's why we still have SCSI" - I would counter that you will see home users accessing data via ADSL, business users via gigabit, and those HPC class systems will have 10 / 40 / 100 gigabit interfaces. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rgb at phy.duke.edu Wed Nov 3 06:55:41 2010 From: rgb at phy.duke.edu (Robert G. Brown) Date: Wed, 3 Nov 2010 06:55:41 -0400 (EDT) Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: On Wed, 3 Nov 2010, John Hearns wrote: > What I think we should be doing is working towards a media-agnostic > form of storing data. Media agnostic with OPEN codices. A great curse on all such IT mechanisms is the closed codex. ASCII works because it is utterly open, as is its logical extension to UTF. But what about music? What about movies? What about books? What about spreadsheets, processed text documents, etc etc? Perhaps html5 will magically solve all such problems, but I doubt it. > A recognition that scientific data (and other forms, like movies and > music etc.) will carry with them metadata and that the data will > migrate through many types of physical media in its lifetime, and will > from the outset have multiple copies made. > I guess the HPC Grid computing types are doing this already, what I'm > rather thinking about is a universal standard for this, and a way of > carrying the metadata with the actual data in a way it cannot be lost. I think this is all dead on correct, but bearing in mind the forces of darkness arrayed on the other side of this, concerned with everything from DRM to encryption to owning and controlling the codex, I personally am not holding my breath. There are also numerous purely technical issues -- modulus problems, for example, in conversion between ogg and mp3 that result in artifacts when switching between lossy compression algorithms that result in nonlinear degradation of information. Similar issues when dealing with old VGA vs 1080p and so on. None of which will go away as the technology evolves. I'm not certain that this is a truly solvable problem. > Its also funny that I use the term "lifetime" - I guess in the past > we all have assumed digital data will have an infinite lifetime, as as > discussed above it has come to pass that the decay of media, or > reading apparatus being unavailable has made data have a finite > lifetime. > > The real point I am making here is that with cloud type data storage > over IP connections even in HPC we will be seeing data accessed not on > SCSI volumes (be that direct SCSI, fibrechannel, iSCSI, RAID etc) but > from an HTTP accessed object store. You might then say that "Hey - > performance matters and that's why we still have SCSI" - I would > counter that you will see home users accessing data via ADSL, business > users via gigabit, and those HPC class systems will have 10 / 40 / 100 > gigabit interfaces. All of which is groovy and I would never argue, but that doesn't address the relative vulnerability of centralized data both to certain kinds of attack and to other kinds of accidents. Or to political control. If Google (ultimately) controls all the data, who controls Google? What happens if they use it for evil instead of for good? How could one stop them from using it for evil if they have your data and also provide you with all of the software you are using to access that data? Who will, after all, guard the guardians? rgb > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From eugen at leitl.org Wed Nov 3 09:37:02 2010 From: eugen at leitl.org (Eugen Leitl) Date: Wed, 3 Nov 2010 14:37:02 +0100 Subject: [Beowulf] Re: Interesting In-Reply-To: References: Message-ID: <20101103133701.GH28998@leitl.org> On Tue, Nov 02, 2010 at 08:49:34PM -0400, Peter St. John wrote: > deBeers laser-engraves serial numbers onto their (natural) diamonds (to > counter the increasing gem quality of artificial diamonds made by, say, > chemical vapor deposition). So how about laser engraving data onto cheap > chemical vapor deposition thin diamond slices? (One of the ideas had been to > make microelectronics substrates from diamond this way, since diamond > conducts heat better than silicon). There's a variant of microfiches which uses photoglass. Areas exposed to light produce crystal nuclei. The glass is heated close to Tg at which stage those nuclei initiate crystal growth. Should have lots longer shelf half life than mere microfiches which are good for about century or more under optimal storage conditions. -- Eugen* Leitl leitl http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From tomislav.maric at gmx.com Wed Nov 3 10:06:27 2010 From: tomislav.maric at gmx.com (tomislav_maric@gmx.com) Date: Wed, 03 Nov 2010 15:06:27 +0100 Subject: [Beowulf] Re: cluster profiling (John Hearns) & (Christopher Samuel) Message-ID: <20101103142415.14270@gmx.com> Thanks a lot for the advice. :) I don't want to overcomplicate things. Its just that I really haven't found any literature that explains how to profile in the top level the cluster and decide where the bottlenecks are. The Ganglia monitoring system has a gstat utility and a python interface that returns the monitoring statistics in a raw mode (not in the HTTP form for its web interface). With the ability to use its Python interface I can write the profiling application within a single language / environment, so this is why I chose it, and the simple way to get the info for each node. I've read online that it has some overhead, but I'm dealing with dedicated multicore nodes and my simulations are really computationally intensive, which will phase out the system side processes on the nodes. I will look up the perf and sysstat definitely in detail before coding anything. Please don't get me wrong, I just want to find a way to build a smal ~20 node multicore COTS beowulf for use with OpenFOAM. This would more than suffice for my needs. In order to do this I don't want to overcomplicate things... I just want to learn as much as I can and build my machine. This kind of profiling harness should inform me on the major metrics for the nodes and for the frontend. I have spent significant amount of time on profiling the application (a solver of the OpenFOAM package) on a single system, only to gradually realize that there are multiple dozens of different CFD/CCM solvers in OpenFOAM, that involve different numerical schemes (spatial, temporal) and operate on different physical fields. This is an impossible task, to optimise the ordinary PC (CPU clock rate, RAM amount, etc) for OpenFOAM, because the number of the parameters involves huge number of permutations. The optimisation itself would take too long and the conclusion is still untouched by the reality: I should just buy the before-last family of processors and RAM, and stack up on the machines. Now, I would just like a few pointers in the right directions regarding the macroscopic metrics of the networked cluster: e.g. increase the simulation size while keeping the core number and the node number constant and find out when the switch drops dead. This will help me evaluate for a specific set of cases (up to 3 Million cells), how much nodes do I need to buy, do I need 10 gig eth, and what kind of speedup did the separation of the network into DATA - HSI - FRONTEND bring me. This also reduces the number of parameters to 1 - node number (in this case 1 or 2, but I'm getting more machines soon) 2 - core number (2 - 8, more to come) 3 - mesh density (scripted increase from 200k to 1 M) the rest are the global metrics: used memory, CPU %, duration of the simulation (user CPU time), network traffic (MB/s) on the switch. The metrics stays the same and is temporaly averaged over the course of the simulation (the case decomposition is static, otherwise, no conclusions may be drawn). I have read the book from Robert Lucke, Robert Gordon Brown, HPC for dummies, the Beowulf books, and a swarm of articles online... and well.... I did my homework... still, I could really use any advice anyone can spare on the profiling/scaling of such machine. Thanks again, Tomislav > ----- Original Message ----- > From: beowulf-request at beowulf.org > Sent: 11/03/10 02:22 AM > To: beowulf at beowulf.org > Subject: Beowulf Digest, Vol 81, Issue 3 > > Send Beowulf mailing list submissions to > beowulf at beowulf.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://www.beowulf.org/mailman/listinfo/beowulf > or, via email, send a message with subject or body 'help' to > beowulf-request at beowulf.org > > You can reach the person managing the list at > beowulf-owner at beowulf.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Beowulf digest..." > > > Today's Topics: > > ?1. cluster profiling (tomislav_maric at gmx.com) > ?2. Re: cluster profiling (John Hearns) > ?3. Re: Re: Interesting (Peter St. John) > ?4. Re: cluster profiling (Christopher Samuel) > ?5. RE: Re: Interesting (Lux, Jim (337C)) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Tue, 02 Nov 2010 22:45:51 +0100 > From: "tomislav_maric at gmx.com" > Subject: [Beowulf] cluster profiling > To: beowulf at beowulf.org > Message-ID: <20101102215301.225020 at gmx.com> > Content-Type: text/plain; charset="utf-8" > > Hi everyone, > > I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. > > The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". > > Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > > Best regards, > Tomislav Maric, (MSc Mechanical Engineering, just to clarify my ignorance regarding HPC) > > > > ------------------------------ > > Message: 2 > Date: Tue, 2 Nov 2010 23:21:13 +0000 > From: John Hearns > Subject: Re: [Beowulf] cluster profiling > To: Beowulf Mailing List > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > > > > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > > > > This sounds very cool. > To be honest, most people use Excel spreadsheets to plot this sort of thing. > If you can produce an automated framework to do this it would be very > interesting. > > I have to slightly question your choice of Ganglia - have you thought > of using sysstat to capture the system's load or memory figures? > > > ------------------------------ > > Message: 3 > Date: Tue, 2 Nov 2010 20:49:34 -0400 > From: "Peter St. John" > Subject: Re: [Beowulf] Re: Interesting > To: "Robert G. Brown" > Cc: "beowulf at beowulf.org" , "Lux, Jim \(337C\)" > > Message-ID: > > Content-Type: text/plain; charset="iso-8859-1" > > deBeers laser-engraves serial numbers onto their (natural) diamonds (to > counter the increasing gem quality of artificial diamonds made by, say, > chemical vapor deposition). So how about laser engraving data onto cheap > chemical vapor deposition thin diamond slices? (One of the ideas had been to > make microelectronics substrates from diamond this way, since diamond > conducts heat better than silicon). > Peter > > On Fri, Oct 29, 2010 at 1:36 PM, Robert G. Brown wrote: > > > On Fri, 29 Oct 2010, Lux, Jim (337C) wrote: > > > > Or, how about something like the UNICON aka "terabit memory" (TBM) from > >> Illiac IV days. It's a stable polyester base with a thin film of rhodium > >> that was ablated by a laser making 3 micron holes to write the bits. > >> $3.5M > >> to store a terabit in 1975. > >> > > > > Burned RO laser disks should in principle be as stable, if the medium > > used is thick enough. The problem is that CDs tend to be mass produced > > with very thin media, cheap plastic, and are even susceptible to > > corrosion through the plastic over time. If one made a CD with tempered > > glass and a moderately thick slice of e.g. stainless steel or > > platinum... > > > > But then your problem is the reader. CD readers give way to DVD and are > > still backwards compatible, sort of. But what about the 2020 > > equivalent? Will there even be one? Nobody will buy actual CDs any > > more. Nobody will buy movies on DVDs any more (seriously, I doubt that > > they will). Will there BE a laser drive that is backwards compatible to > > CD, or will it go the way of reel to reel tapes, 8 track tapes, cassette > > tapes, QIC tapes, floppy drives of all flavors (including high capacity > > drives like the ones I have carefully saved at home in case I ever need > > one), magnetic core memories, large mountable disk packs, exabyte tape > > drives, DA tapes, and so on? I rather think it will be gone. It isn't > > even clear if hard disk drives will still be available (not that any > > computer around would be able to interface with the 5 or 10 MB drives of > > my youth anyway). > > > > This is the problem with electronics. You have to have BOTH long time > > scale stability AND an interface for the ages. And the latter is highly > > incompatible with e.g. Moore's Law -- not even the humble serial port > > has made it through thirty years unscathed. Is the Universal Serial Bus > > really Universal? I doubt it. And yet, that is likely to be the only > > interface available AT ALL (except for perhaps some sort of wireless > > network that isn't even VISIBLE to old peripherals) on the vast bulk of > > the machines sold in a mere five years. > > > > A frightening trend in computing these days is that we may be peaking in > > the era where one's computer (properly equipped with a sensible > > operating system) is symmetrically capable of functioning as a client > > and a server. Desktop computers were clients, servers, or both as one > > wished, from the days of Sun workstations through to the present, with > > any sort of Unixoid operating system and adequate resources. From the > > mid 90's on, with Linux, pure commodity systems were both at the whim of > > the system owner -- anybody could add more memory, more disks, a backup > > device, and the same chassis was whatever you needed it to be. > > > > Now, however, this general purpose desktop is all but dead, supplanted > > by laptops that are just as powerful, but that lack the expandability > > and repurposeability. And laptops are themselves an endangered species > > all of a sudden -- in five years a "laptop" could very well be a single > > "pad" (touchscreen) of whatever size with or without an external > > keyboard, all wireless, smooth as a baby's bottom as far as actual plugs > > are concerned (or maybe, just maybe, with a single USB charger/data port > > or a couple of slots for SD-of-the-day or USB peripherals). Actual data > > storage may well migrate into servers that are completely different > > beasts, far away, accessible only over a wireless network, and > > controlled by others. > > > > An enormous step backwards, in other words. A risk to our political > > freedom. And yet so seductive, so economical, so convenient, that we > > may willingly dance down a primrose path to an information catastrophe > > that is more or less impossible still with the vast decentralization of > > stored knowledge. > > > > rgb > > > > > > Robert G. Brown http://www.phy.duke.edu/~rgb/ > > Duke University Dept. of Physics, Box 90305 > > Durham, N.C. 27708-0305 > > Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu > > > > > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > > http://www.beowulf.org/mailman/listinfo/beowulf > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://www.scyld.com/pipermail/beowulf/attachments/20101102/09fdea41/attachment-0001.html > > ------------------------------ > > Message: 4 > Date: Wed, 03 Nov 2010 11:49:59 +1100 > From: Christopher Samuel > Subject: Re: [Beowulf] cluster profiling > To: beowulf at beowulf.org > Message-ID: <4CD0B1B7.9020705 at unimelb.edu.au> > Content-Type: text/plain; charset=UTF-8 > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On 03/11/10 08:45, tomislav_maric at gmx.com wrote: > > > that will use the Ganglia-python interface and try to > > give me an insight into the way machine is burdened > > during runs > > Depending on how old your kernel is the "perf" utility > (found in the tools/perf directory in your kernel sources, > or packaged in Ubuntu as part of the linux-tools package > or linux-tools-2.6 in Debian Squeeze) may well give you > some interesting stats. > > As a here is an overview of stats a "find -ls" over > the current kernel git tree: > > $ perf stat find . -ls > /dev/null > > ?Performance counter stats for 'find . -ls': > > ?372.415331 task-clock-msecs # 0.923 CPUs > ?158 context-switches # 0.000 M/sec > ?2 CPU-migrations # 0.000 M/sec > ?395 page-faults # 0.001 M/sec > ?648855865 cycles # 1742.291 M/sec > ?698863597 instructions # 1.077 IPC > ?14321645 cache-references # 38.456 M/sec > ?379109 cache-misses # 1.018 M/sec > > ?0.403454703 seconds time elapsed > > > You can use the "perf list" command to get a list of all > the kernel tracepoints you can monitor and then you can > select them individually with the "stat" command. > > Here is perf monitoring CPU migrations, L1 dcache misses > and the kernel scheduler stats of that well known HPC > program "top". ;-) > > perf stat -e migrations -e L1-dcache-load-misses -e sched:* top > > [...] > > ?Performance counter stats for 'top': > > ?0 CPU-migrations # 0.000 M/sec > ?1038307 L1-dcache-load-misses # 0.000 M/sec > ?0 sched:sched_kthread_stop # 0.000 M/sec > ?0 sched:sched_kthread_stop_ret # 0.000 M/sec > ?0 sched:sched_wait_task # 0.000 M/sec > ?98 sched:sched_wakeup # 0.000 M/sec > ?0 sched:sched_wakeup_new # 0.000 M/sec > ?61 sched:sched_switch # 0.000 M/sec > ?0 sched:sched_migrate_task # 0.000 M/sec > ?0 sched:sched_process_free # 0.000 M/sec > ?1 sched:sched_process_exit # 0.000 M/sec > ?0 sched:sched_process_wait # 0.000 M/sec > ?0 sched:sched_process_fork # 0.000 M/sec > ?15 sched:sched_signal_send # 0.000 M/sec > ?49 sched:sched_stat_wait # 0.000 M/sec > ?174 sched:sched_stat_runtime # 0.000 M/sec > ?67 sched:sched_stat_sleep # 0.000 M/sec > ?0 sched:sched_stat_iowait # 0.000 M/sec > > ?29.452075124 seconds time elapsed > > With root access you can even do "perf top" to see what's > going on under the hood. > > You can also use "perf record -g $COMMAND" to record the profiling > information for $COMMAND to perf.data along with call graph information > so you can display a detailed tree view of what was going on via the > "perf report" command. > > Quite a neat little tool I've got to say! > > cheers, > Chris > - -- > ?Christopher Samuel - Senior Systems Administrator > ?VLSCI - Victorian Life Sciences Computational Initiative > ?Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 > ?http://www.vlsci.unimelb.edu.au/ > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.10 (GNU/Linux) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iEYEARECAAYFAkzQsbcACgkQO2KABBYQAh+6JACaAx7p0zARcGGO4busVv7AbqHL > tCcAnA4Z6HOs1LTbucprnyBJFxF6glo+ > =D2wX > -----END PGP SIGNATURE----- > > > ------------------------------ > > Message: 5 > Date: Tue, 2 Nov 2010 18:19:22 -0700 > From: "Lux, Jim (337C)" > Subject: RE: [Beowulf] Re: Interesting > To: "Peter St. John" > Cc: "beowulf at beowulf.org" > Message-ID: > > > Content-Type: text/plain; charset="us-ascii" > > Diamonds may be almost forever, > But I would think that fused silica would work almost as well, and is substantially less expensive. > If you want something exotic, how about ion implantation of Cr+ or Ti+ ions into alumina > > Jim Lux > +1(818)354-2075 > From: Peter St. John [mailto:peter.st.john at gmail.com] > Sent: Tuesday, November 02, 2010 5:50 PM > To: Robert G. Brown > Cc: Lux, Jim (337C); beowulf at beowulf.org > Subject: Re: [Beowulf] Re: Interesting > > deBeers laser-engraves serial numbers onto their (natural) diamonds (to counter the increasing gem quality of artificial diamonds made by, say, chemical vapor deposition). So how about laser engraving data onto cheap chemical vapor deposition thin diamond slices? (One of the ideas had been to make microelectronics substrates from diamond this way, since diamond conducts heat better than silicon). > Peter > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://www.scyld.com/pipermail/beowulf/attachments/20101102/370656f6/attachment.html > > ------------------------------ > > _______________________________________________ > Beowulf mailing list > Beowulf at beowulf.org > http://www.beowulf.org/mailman/listinfo/beowulf > > > End of Beowulf Digest, Vol 81, Issue 3 > ************************************** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Wed Nov 3 10:49:31 2010 From: prentice at ias.edu (Prentice Bisbal) Date: Wed, 03 Nov 2010 10:49:31 -0400 Subject: [Beowulf] cluster profiling In-Reply-To: <20101102215301.225020@gmx.com> References: <20101102215301.225020@gmx.com> Message-ID: <4CD1767B.8090701@ias.edu> tomislav_maric at gmx.com wrote: > Hi everyone, > > I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. > > The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". > > Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > Have you looked at something like Vampir for MPI profiling? Support for VampirTrace is built into OpenMPI, if you compile Open MPI wih the correct options. The rub is that I think you need to pay for a Vampir GUI to analyze the data. I've never used it myself, but I saw a demo once, and it looked pretty powerful. http://www.vampir.eu/ You might also want to look at Tau, PAPI, and Perfmon2 http://www.cs.uoregon.edu/research/tau/home.php http://icl.cs.utk.edu/papi/ http://perfmon2.sourceforge.net/ I set this up for one of my users a couple of years ago. I could be wrong, but I think Tau requires PAPI, and PAPI in turn requires the perfmon2 kernel patches. I could be wrong, since it's been a couple of years. Reading the docs above should point you in the correct direction. That's probably more than you wanted to know. -- Prentice _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From j.wender at science-computing.de Wed Nov 3 11:36:57 2010 From: j.wender at science-computing.de (Jan Wender) Date: Wed, 03 Nov 2010 16:36:57 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: <4CD1767B.8090701@ias.edu> References: <20101102215301.225020@gmx.com> <4CD1767B.8090701@ias.edu> Message-ID: <4CD18199.2030807@science-computing.de> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi all, On 11/03/2010 03:49 PM, Prentice Bisbal wrote: > http://www.vampir.eu/ for some time now (maybe start of 2010) Vampir is not available on its own any more. It got included in intels Cluster Toolkit and is available from them (and underwent a name change if I remember correctly). Cheerio, Jan Wender - -- Jan Wender j.wender at science-computing.de Senior IT Consultant Vertrieb science + computing ag A Bull Group company Hagellocher Weg 73 phone +49 7071 9457 257 72070 Tuebingen, Germany fax +49 7071 9457 522 www.science-computing.de - ---- Company Information ------------------------------------------------------- Vorstand/Board of Management: Dr. Bernd Finkbeiner, Dr. Roland Niemeier, Dr. Arno Steitz, Dr. Ingrid Zech Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Michel Lepert Sitz/Registered Office: Tuebingen Registergericht/Registration Court: Stuttgart Registernummer/Commercial Register No.: HRB 382196 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkzRgZkACgkQXxf+IhbDzz4K1wCeJHKPUcKXszndDB/qeIl7G+4x rrMAn2fzJVxK3Whq/wd2FGKvkAo0VISm =A9mv -----END PGP SIGNATURE----- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From worringen at googlemail.com Wed Nov 3 13:06:25 2010 From: worringen at googlemail.com (Joachim Worringen) Date: Wed, 3 Nov 2010 18:06:25 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: References: <20101102215301.225020@gmx.com> Message-ID: On Wed, Nov 3, 2010 at 12:21 AM, John Hearns wrote: > On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > >> >> Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". >> > > This sounds very cool. > To be honest, most people use Excel spreadsheets to plot this sort of thing. > If you can produce an automated framework to do this it would be very > interesting. Be sure to check out http://perfbase.tigris.org, which is a Python/PostgreSQL toolkit for exactly this purpose. It saves you the nightmare of managing, analysing and visualizing experiment data this crude, manual way. I use it (and designed, write and maintain it) for 5 years now, currently with application to my job in the financial sector. My database currently has some 400GB of data in it. Release 1.2.0 was just released yesterday, which included parallized queries to speed them up. Joachim _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From worringen at googlemail.com Thu Nov 4 04:37:30 2010 From: worringen at googlemail.com (Joachim Worringen) Date: Thu, 4 Nov 2010 09:37:30 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: References: <20101102215301.225020@gmx.com> Message-ID: On Wed, Nov 3, 2010 at 12:21 AM, John Hearns wrote: > On 2 November 2010 21:45, tomislav_maric at gmx.com wrote: > >> >> Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". >> > > This sounds very cool. > To be honest, most people use Excel spreadsheets to plot this sort of thing. > If you can produce an automated framework to do this it would be very > interesting. This is exactly why I wrote "perfbase" some years ago (see http://perfbase.tigris.org), a toolkit for experiment managment and data analysis. I still maintain it and use it on a daily basis (release 1.2.0 came out this week). It parses benchmark output files "automatically", stores data in a database and lets you create and run very elaborate queries on it, generating gnuplots, OpenOffice spreadsheets, raw text or XML output. Joachim _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From tomislav.maric at gmx.com Sat Nov 6 03:37:41 2010 From: tomislav.maric at gmx.com (tomislav_maric@gmx.com) Date: Sat, 06 Nov 2010 08:37:41 +0100 Subject: [Beowulf] Re: cluster profiling (Prentice Bisbal Message-ID: <20101106073741.14260@gmx.com> Thank you very much for the advice. I think that I will try to get the coarse picture of the system load before going into the MPI profiling of the application itself. Since no processes (intensive ones) will run during the testing, I believe it will give me an overview of the system status and let me see where did the bottleneck appear. After this is finished, I will look ino the MPI profiling of the app itself. Thank you again, Tomislav > ----- Original Message ----- > From: beowulf-request at beowulf.org > Sent: 11/03/10 08:00 PM > To: beowulf at beowulf.org > Subject: Beowulf Digest, Vol 81, Issue 5 > > Send Beowulf mailing list submissions to > beowulf at beowulf.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://www.beowulf.org/mailman/listinfo/beowulf > or, via email, send a message with subject or body 'help' to > beowulf-request at beowulf.org > > You can reach the person managing the list at > beowulf-owner at beowulf.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Beowulf digest..." > > > Today's Topics: > > ?1. Re: cluster profiling (Prentice Bisbal) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 03 Nov 2010 10:49:31 -0400 > From: Prentice Bisbal > Subject: Re: [Beowulf] cluster profiling > To: Beowulf Mailing List > Message-ID: <4CD1767B.8090701 at ias.edu> > Content-Type: text/plain; charset=UTF-8 > > > > tomislav_maric at gmx.com wrote: > > Hi everyone, > > > > I'm running a COTS beowlulf cluster and I'm using it for CFD simulations with the OpenFOAM code. I'm currently writing a profiling application (a bunch of scripts) in Python that will use the Ganglia-python interface and try to give me an insight into the way machine is burdened during runs. What I'm actually trying to do is to profile the parallel runs of the OpenFOAM solvers. > > > > The app will increment the mesh density (the coarsness) of the simulation, and run the simulations increasing the number of cores. Right now the machine is miniscule: two nodes with Quad cores. The app will store the data (timing of the execution, the number of cores) and I will plot the diagrams to see when the case size and the core number is starting to drive the speedup away from the "linear one". > > > > Is this a good approach? I know that this will show just tendencies on such an impossible small number of nodes, but I will expand the machine soon, and then their increased number should make these tendencies more accurate. When I cross-reference the temporal data with the system status data given by the ganglia, I can derive conclusions like "O.K., the speedup went down because for the larger cases, the decomposition on max core number was more local, so the system bus must have been burdened, if ganglia confirms that the network is not being strangled for this case configuration". > > > > Can anyone here tell me if I am at least stepping in the right direction? :) Please, don't say "it depends". > > > > Have you looked at something like Vampir for MPI profiling? Support for > VampirTrace is built into OpenMPI, if you compile Open MPI wih the > correct options. > > The rub is that I think you need to pay for a Vampir GUI to analyze the > data. I've never used it myself, but I saw a demo once, and it looked > pretty powerful. > > http://www.vampir.eu/ > > You might also want to look at Tau, PAPI, and Perfmon2 > > http://www.cs.uoregon.edu/research/tau/home.php > http://icl.cs.utk.edu/papi/ > http://perfmon2.sourceforge.net/ > > I set this up for one of my users a couple of years ago. I could be > wrong, but I think Tau requires PAPI, and PAPI in turn requires the > perfmon2 kernel patches. I could be wrong, since it's been a couple of > years. Reading the docs above should point you in the correct direction. > > That's probably more than you wanted to know. > > > > -- > Prentice > > > ------------------------------ > > _______________________________________________ > Beowulf mailing list > Beowulf at beowulf.org > http://www.beowulf.org/mailman/listinfo/beowulf > > > End of Beowulf Digest, Vol 81, Issue 5 > ************************************** _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From dnlombar at ichips.intel.com Mon Nov 8 16:34:23 2010 From: dnlombar at ichips.intel.com (David N. Lombard) Date: Mon, 8 Nov 2010 13:34:23 -0800 Subject: [Beowulf] cluster profiling In-Reply-To: <4CD18199.2030807@science-computing.de> References: <20101102215301.225020@gmx.com> <4CD1767B.8090701@ias.edu> <4CD18199.2030807@science-computing.de> Message-ID: <20101108213422.GA5621@nlxcldnl2.cl.intel.com> On Wed, Nov 03, 2010 at 08:36:57AM -0700, Jan Wender wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi all, > > On 11/03/2010 03:49 PM, Prentice Bisbal wrote: > > http://www.vampir.eu/ > > for some time now (maybe start of 2010) Vampir is not available on its own any > more. It got included in intels Cluster Toolkit and is available from them (and > underwent a name change if I remember correctly). The Intel Trace Analyzer and Collector (ITAC), successor of Pallas Vampir, is available in the Intel Cluster Toolkit. But, that's not a recent event, Intel acquired the HPC group of Pallas in 2003. BTW, the Intel MPI Benchmarks, ex Pallas MPI Benchmarks, are similarly available. -- David N. Lombard, Intel, Irvine, CA I do not speak for Intel Corporation; all comments are strictly my own. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Wed Nov 17 14:08:19 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Wed, 17 Nov 2010 14:08:19 -0500 Subject: [Beowulf] LAPACK C/C++ wrappers Message-ID: I'm going to dabble a bit with writing some C/C++ with a linear algebra library. My fortran is rusty (although I did some work translating fortran to C in this millennium :-). The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice have any real bearing on compiling against MPI later? Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) I'd be glad to hear it. I just installed (k)ubuntu, which seems fine for my purposes so far. I do miss fvwm, and the default "single click to invoke an application" is totally nuts but I fixed it. Thanks, Peter -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From ispmarin at gmail.com Sun Nov 21 23:55:02 2010 From: ispmarin at gmail.com (Ivan Marin) Date: Sun, 21 Nov 2010 22:55:02 -0600 Subject: [Beowulf] LAPACK C/C++ wrappers In-Reply-To: References: Message-ID: I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I did some tests in the past for simple stuff, and worked, but very simple linear system solvers. Ivan Marin Civil Engineering Dept University of Minnesota 500 Pillsbury Dr. S.E. Minneapolis, MN 55455 Laborat?rio de Hidr?ulica Computacional - LHC Departamento de Hidr?ulica e Saneamento - SHS Escola de Engenharia de S?o Carlos - EESC Universidade de S?o Paulo - USP http://albatroz.shs.eesc.usp.br +55 16 3373 8270 2010/11/17 Peter St. John : > I'm going to dabble a bit with writing some C/C++ with a linear algebra > library. My fortran is rusty (although I did some work translating fortran > to C in this millennium :-). > The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice have > any real bearing on compiling against MPI later? > Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) I'd > be glad to hear it. > > I just installed (k)ubuntu, which seems fine for my purposes so far. I do > miss fvwm, and the default "single click to invoke an application" is > totally nuts but I fixed it. > Thanks, > Peter > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From j.wender at science-computing.de Mon Nov 22 04:29:33 2010 From: j.wender at science-computing.de (Jan Wender) Date: Mon, 22 Nov 2010 10:29:33 +0100 Subject: [Beowulf] cluster profiling In-Reply-To: <4CD18199.2030807@science-computing.de> References: <20101102215301.225020@gmx.com> <4CD1767B.8090701@ias.edu> <4CD18199.2030807@science-computing.de> Message-ID: <20101122092932.GA28804@sanja.science-computing.de> Hi, I have to correct myself... On Wed, Nov 03, 2010 at 04:36:57PM +0100, Jan Wender wrote: > for some time now (maybe start of 2010) Vampir is not available on its own any > more. It got included in intels Cluster Toolkit and is available from them (and > underwent a name change if I remember correctly). At SC10 I found the Vampir booth and had a short chat with them. Their tool was included at one time with the intel Cluster tools, but they did not stop to develop it further on their own. You can get their version on www.vampir.eu and http://tu-dresden.de/die_tu_dresden/zentrale_einrichtungen/zih/forschung/software_werkzeuge_zur_unterstuetzung_von_programmierung_und_optimierung/ Cheerio, Jan -- Jan Wender j.wender at science-computing.de Senior IT Consultant Vertrieb science + computing ag A Bull Group Company Hagellocher Weg 73 phone +49 7071 9457 257 72070 Tuebingen, Germany fax +49 7071 9457 522 www.science-computing.de ---- Company Information ------------------------------------------------------- Vorstand/Board of Management: Dr. Bernd Finkbeiner, Dr. Roland Niemeier, Dr. Arno Steitz, Dr. Ingrid Zech Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Michel Lepert Sitz/Registered Office: Tuebingen Registergericht/Registration Court: Stuttgart Registernummer/Commercial Register No.: HRB 382196 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: not available URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From eugen at leitl.org Mon Nov 22 10:50:07 2010 From: eugen at leitl.org (Eugen Leitl) Date: Mon, 22 Nov 2010 16:50:07 +0100 Subject: [Beowulf] Intel: 1,000-core processor possible Message-ID: <20101122155007.GI9434@leitl.org> http://www.goodgearguide.com.au/article/368762/intel_1_000-core_processor_possible/ Intel: 1,000-core processor possible A group of Intel researchers has pioneered a messaging system that would allow multiple cores to communicate Joab Jackson (IDG News Service)20 November, 2010 09:04 An experimental Intel chip shows the feasibility of building processors with 1,000 cores, an Intel researcher has asserted. The architecture for the Intel 48-core Single Chip Cloud Computer[1] (SCC) processor is "arbitrarily scalable," said Intel researcher Timothy Mattson, during a talk at the Supercomputer 2010 conference being held this week in New Orleans. "This is an architecture that could, in principle, scale to 1,000 cores," he said. " I can just keep adding, adding, adding cores." Only after 1,000 cores or so, the diameter of the mesh, or the on-chip network connecting the many cores, will grow to such an extent that it would negatively impact performance, Mattson said. Intel remains adamant that the future progress of microprocessors will depend on packing ever more cores onto a chip. As more cores are added, however, Intel designers must confront the problem of scalability. Initial multicore chip architectures depended on a set of protocols that assures that each core has the same view of the system's memory, a technique called cache coherency. As more cores are added to chips, this approach becomes problematic insofar that "the protocol overhead per core grows with the number of cores, leading to a 'coherency wall' beyond which the overhead exceeds the value of adding cores," the paper accompanying Mattson's talk noted. Mattson has argued[2] that a better approach would be to eliminate cache coherency and instead allow cores to pass messages among one another. The recent work of the design team has centered on developing message-passing techniques for the chip that would scale as more cores are added. Designed by Intel's TeraScale Research Program over the past several years, the chip itself is an experimental one and is not on the Intel product road map, Mattson said. A limited number of copies have been distributed to researchers and developers so they can build development tools for the design. The chip, first fabricated with a 45-nanometer process at Intel facilities about a year ago, is actually a six-by-four array of tiles, each tile containing two cores. It has more than 1.3 billion transistors and consumes from 25 to 125 watts. For simplicity's sake, the team used an off-the-shelf 1994-era Pentium processor design for the cores themselves. "Performance on this chip is not interesting," Mattson said. It uses a standard x86 instruction set. The novelty of this processor is in its tiled architecture and the network and address infrastructure. Each core has a "mesh interface component" that packages data into packets and connects to an on-board router. Each tile also has a "message-passing buffer," with 16 kilobytes of random access memory. The team has tried various approaches to streamline the ability of the processor to pass messages among the many cores. By installing the TCP/IP protocol on the data link layer, the team was able to run a separate Linux-based operating system on each core. Mattson noted that while it would be possible to run a 48-node Linux cluster on the chip, it "would be boring." "To make this interesting, I would have to ask, how would the programming models map onto the unique features of this chip," he said. The team also developed a small API (application programming interface) library for message passing among the cores, called RCCE, and which Mattson pronounced as "Rocky." In tests, the team showed that message passing among the cores could be just as speedy using RCCE as with TCP/IP-based Linux cluster. And both approaches bode well for the message-passing approach for inter-core communication. "Our preliminary work has demonstrated that the SCC processor and its native message passing API provide an effective software development platform," the paper concludes. "The expected difficulties due to the lack of asynchronous message passing have so far not materialized." In addition to talking about the chip's message-passing capabilities, Mattson also elaborated on SCC's power-saving capabilities[3]. The frequency of each tile can be varied. Hooks are provided for programmers that would allow their programs to adjust the frequency speed and even the voltage of the cores they are running upon. This feature will, however, create a new challenge for programmers, he warned. "It's a lot harder than you'd think to look at your program and think 'how many volts do I really need?'" he said. Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson[4]. Joab's e-mail address is Joab_Jackson at idg.com[5] References http://www.computerworld.com/s/article/9174981/Intel_to_ship_samples_of_experimental_48_core_chip http://og-hpc.com/Rice2010/Slides/Mattson-OG-HPC-2010-Intel.pdf http://www.pcworld.com/businesscenter/article/183653/intel_48core_singlechip_cloud_computer_improves_power_efficiency.html http://twitter.com/Joab_Jackson mailto:Joab_Jackson at idg.com _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From gus at ldeo.columbia.edu Mon Nov 22 11:42:42 2010 From: gus at ldeo.columbia.edu (Gus Correa) Date: Mon, 22 Nov 2010 11:42:42 -0500 Subject: [Beowulf] (no subject) In-Reply-To: References: Message-ID: <4CEA9D82.3020408@ldeo.columbia.edu> Would PETSc be of use for you? http://www.mcs.anl.gov/petsc/petsc-as/ Gus Correa Ivan Marin wrote: > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I > did some tests in the past for simple stuff, and worked, but very > simple linear system solvers. > > Ivan Marin > > Civil Engineering Dept > University of Minnesota > 500 Pillsbury Dr. S.E. > Minneapolis, MN 55455 > > Laborat?rio de Hidr?ulica Computacional - LHC > Departamento de Hidr?ulica e Saneamento - SHS > Escola de Engenharia de S?o Carlos - EESC > Universidade de S?o Paulo - USP > > http://albatroz.shs.eesc.usp.br > +55 16 3373 8270 > > > > 2010/11/17 Peter St. John : >> I'm going to dabble a bit with writing some C/C++ with a linear algebra >> library. My fortran is rusty (although I did some work translating fortran >> to C in this millennium :-). >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice have >> any real bearing on compiling against MPI later? >> Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) I'd >> be glad to hear it. >> >> I just installed (k)ubuntu, which seems fine for my purposes so far. I do >> miss fvwm, and the default "single click to invoke an application" is >> totally nuts but I fixed it. >> Thanks, >> Peter >> >> _______________________________________________ >> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing >> To change your subscription (digest mode or unsubscribe) visit >> http://www.beowulf.org/mailman/listinfo/beowulf >> >> > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Mon Nov 22 12:02:20 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Mon, 22 Nov 2010 12:02:20 -0500 Subject: [Beowulf] ("LAPACK/C++ wrappers) Message-ID: Gus, In my case, I"m not concerned with PDEs but the "linear system solvers" themselves,which PETSci gets from other packages? So for example, looking over their table at: http://www.mcs.anl.gov/petsc/petsc-as/documentation/linearsolvertable.html I'd want the "dense" case and therefore PLAPACK, so now I know to google PLAPACK. I just want to solve a system of simultaneous linear equations within the vernacular of C/C++. Thanks, Peter On Mon, Nov 22, 2010 at 11:42 AM, Gus Correa wrote: > Would PETSc be of use for you? > > http://www.mcs.anl.gov/petsc/petsc-as/ > > Gus Correa > > Ivan Marin wrote: > > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I > > did some tests in the past for simple stuff, and worked, but very > > simple linear system solvers. > > > > Ivan Marin > > > > Civil Engineering Dept > > University of Minnesota > > 500 Pillsbury Dr. S.E. > > Minneapolis, MN 55455 > > > > Laborat?rio de Hidr?ulica Computacional - LHC > > Departamento de Hidr?ulica e Saneamento - SHS > > Escola de Engenharia de S?o Carlos - EESC > > Universidade de S?o Paulo - USP > > > > http://albatroz.shs.eesc.usp.br > > +55 16 3373 8270 > > > > > > > > 2010/11/17 Peter St. John : > >> I'm going to dabble a bit with writing some C/C++ with a linear algebra > >> library. My fortran is rusty (although I did some work translating > fortran > >> to C in this millennium :-). > >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this choice > have > >> any real bearing on compiling against MPI later? > >> Just if anyone is happy with any combination (LAPACK++ and OpenMPI, say) > I'd > >> be glad to hear it. > >> > >> I just installed (k)ubuntu, which seems fine for my purposes so far. I > do > >> miss fvwm, and the default "single click to invoke an application" is > >> totally nuts but I fixed it. > >> Thanks, > >> Peter > >> > >> _______________________________________________ > >> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin > Computing > >> To change your subscription (digest mode or unsubscribe) visit > >> http://www.beowulf.org/mailman/listinfo/beowulf > >> > >> > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From gus at ldeo.columbia.edu Mon Nov 22 12:38:19 2010 From: gus at ldeo.columbia.edu (Gus Correa) Date: Mon, 22 Nov 2010 12:38:19 -0500 Subject: [Beowulf] ("LAPACK/C++ wrappers) In-Reply-To: References: Message-ID: <4CEAAA8B.8060702@ldeo.columbia.edu> Peter St. John wrote: > Gus, > In my case, I"m not concerned with PDEs but the "linear system solvers" > themselves,which PETSci gets from other packages? Hi Peter Yes, you can compile PETSC with support of a variety of linear algebra packages. See their installation instructions and documentation. I think there is support for PLAPACK, please check their install and docs pages: http://www.mcs.anl.gov/petsc/petsc-2/documentation/installation.html http://www.mcs.anl.gov/petsc/petsc-2/documentation/index.html > So for example, > looking over their table at: > http://www.mcs.anl.gov/petsc/petsc-as/documentation/linearsolvertable.html > I'd want the "dense" case and therefore PLAPACK, so now I know to > google PLAPACK. > I just want to solve a system of simultaneous linear equations within > the vernacular of C/C++. AFAIK, PETSc has C, Fortran, and C++ interfaces. If I remember right, you can choose between C++ or C when you compile it. You can choose between real (for most linear algebra ) or complex (a must for FFTs, and FFTW support) scalars. You can also choose which linear algebra packages you will use (which you can compile separately, or let PETSc handle - it handles many of them), which BLAS/LAPACK to use (say, Goto BLAS, Intel MKL, AMD ACML, or let PETSc install its own version). I am not a PETSc user, but some people here used PETSc to solve inverse and forward problems, problems with sparse and dense matrices, etc. As it often happens when science (not computer science) is the main goal, once the "it works" stage is reached, the code starts to be used to produce papers, more and more features are added, and it is never again improved for efficiency, never goes beyond the prototype phase in this regard. In terms of efficiency, the results didn't shine, in terms of science/results output they were very good. However, it may well be that the way the problem (not PETSc) was programmed led to the inefficient code. PETSc experts in the list: For the benefit of Peter, Ivan, and the list, would you kindly jump in and say something about PETSc efficiency and other features? Txs. Maybe PETSc is an overkill for you, but if the problem is that simple, why not use Fortran and regular LAPACK/PLAPACK? My two cents. Gus Correa > Thanks, > Peter > > On Mon, Nov 22, 2010 at 11:42 AM, Gus Correa > wrote: > > Would PETSc be of use for you? > > http://www.mcs.anl.gov/petsc/petsc-as/ > > Gus Correa > > Ivan Marin wrote: > > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. I > > did some tests in the past for simple stuff, and worked, but very > > simple linear system solvers. > > > > Ivan Marin > > > > Civil Engineering Dept > > University of Minnesota > > 500 Pillsbury Dr. S.E. > > Minneapolis, MN 55455 > > > > Laborat?rio de Hidr?ulica Computacional - LHC > > Departamento de Hidr?ulica e Saneamento - SHS > > Escola de Engenharia de S?o Carlos - EESC > > Universidade de S?o Paulo - USP > > > > http://albatroz.shs.eesc.usp.br > > +55 16 3373 8270 > > > > > > > > 2010/11/17 Peter St. John >: > >> I'm going to dabble a bit with writing some C/C++ with a linear > algebra > >> library. My fortran is rusty (although I did some work > translating fortran > >> to C in this millennium :-). > >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this > choice have > >> any real bearing on compiling against MPI later? > >> Just if anyone is happy with any combination (LAPACK++ and > OpenMPI, say) I'd > >> be glad to hear it. > >> > >> I just installed (k)ubuntu, which seems fine for my purposes so > far. I do > >> miss fvwm, and the default "single click to invoke an > application" is > >> totally nuts but I fixed it. > >> Thanks, > >> Peter > >> > >> _______________________________________________ > >> Beowulf mailing list, Beowulf at beowulf.org > sponsored by Penguin Computing > >> To change your subscription (digest mode or unsubscribe) visit > >> http://www.beowulf.org/mailman/listinfo/beowulf > >> > >> > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org > sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org > sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From peter.st.john at gmail.com Mon Nov 22 13:35:10 2010 From: peter.st.john at gmail.com (Peter St. John) Date: Mon, 22 Nov 2010 13:35:10 -0500 Subject: [Beowulf] ("LAPACK/C++ wrappers) In-Reply-To: <4CEAAA8B.8060702@ldeo.columbia.edu> References: <4CEAAA8B.8060702@ldeo.columbia.edu> Message-ID: Regarding "...Maybe PETSc is an overkill for you, but if the problem is that simple, why not use Fortran and regular LAPACK/PLAPACK?,,," I haven't written fortran since IV and 77 in the early 80's. The matrix multiplication part is simple but the ambient application is squirrelly and I"m more facile in C or C++. So I just want basic linear algebra library I can link my C to, but maybe all y'all use Matlab for simple things. Probably I should just do it with, say, Armadillo, and learn something without spilling too much blood from fingertips. Peter On Mon, Nov 22, 2010 at 12:38 PM, Gus Correa wrote: > Peter St. John wrote: > > Gus, > > In my case, I"m not concerned with PDEs but the "linear system solvers" > > themselves,which PETSci gets from other packages? > Hi Peter > > Yes, you can compile PETSC with support of a variety of linear algebra > packages. See their installation instructions and documentation. > I think there is support for PLAPACK, please check their > install and docs pages: > > http://www.mcs.anl.gov/petsc/petsc-2/documentation/installation.html > http://www.mcs.anl.gov/petsc/petsc-2/documentation/index.html > > > So for example, > > looking over their table at: > > > http://www.mcs.anl.gov/petsc/petsc-as/documentation/linearsolvertable.html > > I'd want the "dense" case and therefore PLAPACK, so now I know to > > google PLAPACK. > > I just want to solve a system of simultaneous linear equations within > > the vernacular of C/C++. > > AFAIK, PETSc has C, Fortran, and C++ interfaces. > If I remember right, you can choose between C++ or C when > you compile it. > You can choose between real (for most linear algebra ) > or complex (a must for FFTs, and FFTW support) scalars. > You can also choose which > linear algebra packages you will use (which you can compile separately, > or let PETSc handle - it handles many of them), > which BLAS/LAPACK to use (say, Goto BLAS, > Intel MKL, AMD ACML, or let PETSc install its own version). > > I am not a PETSc user, but some people here used PETSc to solve inverse > and forward problems, problems with sparse and dense matrices, etc. > As it often happens when science (not computer science) > is the main goal, once the "it works" stage is reached, > the code starts to be used to produce papers, > more and more features are added, > and it is never again improved for efficiency, > never goes beyond the prototype phase in this regard. > > In terms of efficiency, the results didn't shine, > in terms of science/results output they were very good. > However, it may well be that the way the problem > (not PETSc) was programmed led to the inefficient code. > > PETSc experts in the list: > For the benefit of Peter, Ivan, and the list, > would you kindly jump in and say something about PETSc > efficiency and other features? > Txs. > > Maybe PETSc is an overkill for you, but if the problem is > that simple, why not use Fortran and regular LAPACK/PLAPACK? > > My two cents. > Gus Correa > > > Thanks, > > Peter > > > > On Mon, Nov 22, 2010 at 11:42 AM, Gus Correa > > wrote: > > > > Would PETSc be of use for you? > > > > http://www.mcs.anl.gov/petsc/petsc-as/ > > > > Gus Correa > > > > Ivan Marin wrote: > > > I'm interested also in the bindings for C++ and LAPACK/ScaLAPACK. > I > > > did some tests in the past for simple stuff, and worked, but very > > > simple linear system solvers. > > > > > > Ivan Marin > > > > > > Civil Engineering Dept > > > University of Minnesota > > > 500 Pillsbury Dr. S.E. > > > Minneapolis, MN 55455 > > > > > > Laborat?rio de Hidr?ulica Computacional - LHC > > > Departamento de Hidr?ulica e Saneamento - SHS > > > Escola de Engenharia de S?o Carlos - EESC > > > Universidade de S?o Paulo - USP > > > > > > http://albatroz.shs.eesc.usp.br > > > +55 16 3373 8270 > > > > > > > > > > > > 2010/11/17 Peter St. John > >: > > >> I'm going to dabble a bit with writing some C/C++ with a linear > > algebra > > >> library. My fortran is rusty (although I did some work > > translating fortran > > >> to C in this millennium :-). > > >> The wiki suggests: LAPACK++, clapack, and Armadillo. Would this > > choice have > > >> any real bearing on compiling against MPI later? > > >> Just if anyone is happy with any combination (LAPACK++ and > > OpenMPI, say) I'd > > >> be glad to hear it. > > >> > > >> I just installed (k)ubuntu, which seems fine for my purposes so > > far. I do > > >> miss fvwm, and the default "single click to invoke an > > application" is > > >> totally nuts but I fixed it. > > >> Thanks, > > >> Peter > > >> > > >> _______________________________________________ > > >> Beowulf mailing list, Beowulf at beowulf.org > > sponsored by Penguin Computing > > >> To change your subscription (digest mode or unsubscribe) visit > > >> http://www.beowulf.org/mailman/listinfo/beowulf > > >> > > >> > > > _______________________________________________ > > > Beowulf mailing list, Beowulf at beowulf.org > > sponsored by Penguin Computing > > > To change your subscription (digest mode or unsubscribe) visit > > http://www.beowulf.org/mailman/listinfo/beowulf > > > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org > > sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > > http://www.beowulf.org/mailman/listinfo/beowulf > > > > > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From ispmarin at gmail.com Tue Nov 23 11:48:37 2010 From: ispmarin at gmail.com (Ivan Marin) Date: Tue, 23 Nov 2010 10:48:37 -0600 Subject: [Beowulf] C/C++ Numerical Integral solvers In-Reply-To: <4CEAAA8B.8060702@ldeo.columbia.edu> References: <4CEAAA8B.8060702@ldeo.columbia.edu> Message-ID: <4CEBF065.4070809@gmail.com> Hi folks, I'm following closely the thread about the numerical solvers in C++, and would like to ask: is anyone using or has experience with numerical integral solvers, hopefully parallel, in C++? I'm solving singular Cauchy type integrals, complex. I've tried the GSL and they are very precise, but _very_ slow for my type of integrands, and not intrinsically threaded/parallel. -- Ivan Marin Civil Engineering Dept University of Minnesota 500 Pillsbury Dr. S.E. Minneapolis, MN 55455 Laborat?rio de Hidr?ulica Computacional - LHC Departamento de Hidr?ulica e Saneamento - SHS Escola de Engenharia de S?o Carlos - EESC Universidade de S?o Paulo - USP http://albatroz.shs.eesc.usp.br +55 16 3373 8270 _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.