Home

Details
Hits: 16502

Note: this paper was prepared for a conference that we decided not to attend (Okay, it was not accepted). It is written in a more formal style than the normal ClusterMonkey articles and is sponsored by the The Beowulf Foundation

Abstract

Popular homogeneous clustered HPC systems (e.g., commodity x86 servers connected by a high-speed interconnect) have given way to heterogeneous clusters comprised of multi-core servers, high speed interconnects, accelerators (often GPU based), and custom storage arrays. Cluster designers are often faced with finding a balance between purpose-built (tailored to specific problem domains ) and general use systems. Traditional cluster-based approaches, however, all share a hard boundary between internal server buses (mainly PCIe) and the rest of the cluster. In heterogeneous environments, the server boundary often creates inefficient resource management, limits solution flexibility, and heavily influences the design of clustered HPC applications. This paper explores the malleability of the GigaIO™ FabreX™ PCIe memory fabric in relation to HPC cluster applications. A discussion of emerging concepts (e.g., a routable PCIe bus) and hands-on benchmarks using shared GPUs will be provided. In addition, results of a simple integration with SLURM resource scheduler will be discussed as way to make composable/malleable computing transparently available to end-users. Keywords. Composable computing, malleable computing, PCIe, HPC cluster, SLURM, benchmark, FabreX , GigaIO, resource scheduler
Read more: First Experiences with Composable Hardware: An HPC User Perspective Add new comment

Details
Hits: 9867

From the What's Watts Dept.

Some Clarity around TDP ratings

Managing power usage on multi-core processors has become an important aspect with modern computing systems. At the same time, finding an accurate specification of actual power usage has become more difficult. Knowing power usage is important in many areas and particularly when considering the efficiency of High Performance Computing (HPC) systems. In almost all modern CPUs and GPUs the only number that seems to give a hint about power usage is Thermal Design Power or TDP.

According to Wikipedia Thermal Design Power is defined as follows:

... is the maximum amount of heat generated by a computer chip or component (often a CPU, GPU or system on a chip) that the cooling system in a computer is designed to dissipate under any workload.

Some sources state that the peak power rating for a microprocessor is usually 1.5 times the TDP rating

Read more: These are Not the Watts You Are Looking For Add new comment

Details
Hits: 3848

From the self reference department

The Beowulf Foundation was announced by Douglas Eadline and Lara Kisielewska at the SC 2021 Beowulf Bash in St Louis. The idea for a “Foundation” emerged from the continued discussion amongst the Beowulf community about the supporting the “Beowulf Ethos,” that began with the Beowulf Project at NASA using commodity hardware and open source software to build high-performance systems at low cost. Initially considered an anomaly, the “wrong ideas” demonstrated by the Beowulf Project changed the face of modern supercomputing.

The goal of the Beowulf Foundation is not to relive the past, but to support the “wrong” ideas of the future that may lead to further breakthroughs in high performance computing. As stated by Eadline:

Read more: Disrupt Forward: Announcing the Beowulf Foundation Add new comment

Details
Hits: 4605

From the "What We Have Been Saying Dept"

Fresh from Julia Computing is a short note about HPC maven Jack Dongarra had to say about Julia.

2021 Turing Award Winner Jack Dongarra Says Julia Is ‘Much Better’ Than Other Languages and Should Perhaps Take Over: The winner of the 2021 Turing Award, often referred to as the ‘Nobel Prize of Computing’, is Jack Dongarra. Dongarra says Julia is ‘much better’ than other languages and should perhaps take over. According to ZDNet:

“While hardware speeds up matrix multiplication, [Jack] Dongarra is, again, mindful of the needs of the scientists and the software writer. ‘I grew up writing FORTRAN, and today we have much better mechanisms’ such as the Julia programming language and Jupyter Notebooks. What's needed now, he said, are more ways to ‘express those computations in an easy way,’ meaning linear algebra computations such as matrix multiplications. Specifically, more tools are needed to abstract the details. ‘Making the scientist more productive is the right way to go,’ he said. Asked what software programming paradigm should perhaps take over, Dongarra suggested the Julia language is one good candidate …”

Of course some forward looking sites, (cough, cough) have been highlighting Julia for ten years,

Add new comment

Details
Hits: 5483
From the "So That is What My Data Look Like" Department

A new open source, portable, and scalable 3D visualization standard is in the works. Announced by Khronos Group ANARI (Analytic Rendering Interface) 1.0 Provisional Specification for scientific visualization and scalable 3D Data visualization. ANARI will enable users to quickly and easily build a scene description and create 3D images. Most importantly, ANARI doesn’t specify the rendering details; the rendering gets left to the backend engine which provides a simplified way to develop a visualization application with cross-vendor portability including ray tracing. From the announcement:

"The scientific visualization ecosystem includes key visualization application vendors such as VMD, VTK/ParaView, and VisIt that have actively participated in ANARI’s design. There are also early ANARI implementations in development by AMD, Intel and NVIDIA that will provide access to their Radeon ProRender, OSPRay, and VisRTX rendering engines respectively".

The specification is provisional and a full version 1 should be released soon (there is an open SDK available). There is also a good (and short) Overview Video that explains the benefits of ANARI.

Add new comment

Main Menu

Search

Login And Newsletter

Create an account to access exclusive content, comment on articles, and receive our newsletters.

Feedburner

Who's Online

We have 68 guests and no members online

feed-image Front Page RSS Feed

This work is licensed under CC BY-NC-SA 4.0

©2005-2023 Copyright Seagrove LLC, Some rights reserved. Except where otherwise noted, this site is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International. The Cluster Monkey Logo and Monkey Character are Trademarks of Seagrove LLC.