[Beowulf] SC|05: will 10Ge beat IB?

Joel Jaeggli joelja at darkwing.uoregon.edu
Tue Nov 22 13:30:05 EST 2005


On Tue, 22 Nov 2005, Mark Hahn wrote:

>> Chelsio's T210-CX4 TOE NIC has been around for about a year. Like the
>> Myricom product it utilizes the 10GBase-CX standard. It essentially
>> pushes 10GbE over InfiniBand cabling directly using the XAUI interface
>> (normally used for on-PCB communication on 10G devices).
>
> indeed, it's CX4 that I was seeing all over the place.  yes, it's a
> pretty heavy/thick cable, and not cheap, and limited to ~15M.
> but for clusters, 15M is pretty workable.
>
>> relatively cheap UTP cabling and are limited to 50ft per link. It seems
>> to me to be a stopgap standard that will be overtaken by 10Gbase-T in a
>> few years time.
>
> is there really a promise of running 10G on something like cat5?

Think cat7, ISO/IEC 11801:2002 category 7/class F, but the answer is yes.

that should get you to at least 600mhz with gg45/rj-45 connectors on 
either end and higher with some alternate connector types.

>> I don't have a quote on the pricing but it should be competitive with
>> IB.
>
> my understanding is that an IB cable has some active equalization
> built into the connectors, and that this is the only difference between
> it and eth-style CX4.  I also understand that putting the equalization
> on the cable size have been a design mistake, since it means that distortion
> caused by the connector itslef is not corrected...
>
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
>

-- 
--------------------------------------------------------------------------
Joel Jaeggli  	       Unix Consulting 	       joelja at darkwing.uoregon.edu
GPG Key Fingerprint:     5C6E 0104 BAF0 40B0 5BD3 C38B F000 35AB B67F 56B2

_______________________________________________
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf



More information about the Beowulf mailing list