Printer Friendly

Research and Markets: High Performance Computing User Site Census: Processors.

DUBLIN -- Research and Markets (http://www.researchandmarkets.com/research/eaa4c7/high_performance_c) has announced the addition of Intersect360 Research's new report "High Performance Computing User Site Census: Processors" to their offering.

The Processors report is part of Intersect360's Site Census series and provides an examination of the processor use in systems found at a sample of HPC user sites. Intersect360 surveyed a broad range of users about their current computer system installations, storage systems, networks, middleware, and applications software supporting these computer installations. The goal in this analysis is to examine the processor suppliers, products, and configurations used in high-performance computing systems.

By three measures of share, the x86 architecture is predominant in the HPC community. Ninety percent of systems in the total site survey reported processors from AMD or Intel. The data shows this same trend when examining processors by node count and by core count. Intel remains the market share leader. All vendors other than Intel and AMD have single-digit shares. This analysis was performed excluding outliers, which are systems with 2,000 or more nodes.

Two-processor nodes remain the most common in distributed memory systems, although 4, 8, and 16 processor nodes do appear in the Census population. The cluster architecture remains dominant with smaller contributions from blade and SMP systems. Some systems have a mix of nodes with 2 and 4 processors, and other combinations.

Multi-core processor technology is increasing common in HPC systems and the number of cores per processor is rising. When studied by year of acquisition, average number of cores per processor has nearly doubled between 2006 and 2009. The four core processor holds the greatest share and is the sweet spot for HPC designers and processor suppliers at this time.

Trends in memory requirements anticipate increasing challenges for the HPC community. Average memory per core remains about the same over the last five years and average memory per processor has doubled. However, the average memory used by nodes is growing non-linearly. Providing sufficient memory to new multi-core processor nodes will be increasingly difficult as this trend continues.

The use of special-purpose processors, or accelerators, remains relatively rare in the HPC community, but growth continues. Nvidia is most often named as the accelerator supplier, but many respondents indicated that they did not know the suppliers name.

For more information visit http://www.researchandmarkets.com/research/eaa4c7/high_performance_c

Key Topics Covered:

EXECUTIVE SUMMARY INTRODUCTION METHODOLOGY

* Survey Methodology

* Methodology Notes

* Demographics

RESULTS AND ANALYSIS

* System, Node, Processor and Core Distributions by Supplier

* AMD and Intel System Share by System Size

* AMD and Intel System Share by Acquisition Year

* System and Processor Distributions by Processor Architecture

* Processors per Node within Distributed Memory Systems

* Cores per Processor

* Systems, Processors, and Cores by System Architecture

* Memory Requirements per Node, Processor and Core

* Accelerators

CONCLUSIONS

* Ongoing Strategies

* Preparing for the Next Big Thing

* Overall Perspective
COPYRIGHT 2011 Business Wire
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Business Wire
Date:Mar 14, 2011
Words:567
Previous Article:Eaton Agrees to Acquire Internormen Technology Group, Adding Additional Capabilities to its Global Filtration Business.
Next Article:Cumulus Reports Fourth Quarter 2010 Results.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters