SpyderByte: WinHPC.org EnterpriseLinux.org BigBlueLinux.org
      
 The #1 Site for News & Information Related to Linux High Performance Technical Computing, Linux High Availability and Linux Parallel Clustering
Home About News Archives Contribute News, Articles, Press Releases Mobile Edition Contact Sponsorship Search Privacy
More Links
  • Printer friendly version
  • Share this story

  • Research and Services
    Cluster Quoter
    Windows HPC News
    Cluster Builder
    Hardware Vendors
    Software Vendors
    Training Vendors
    Golden Eggs
    Forums
    Employment/Jobs
    Beowulf
    Applications
    Interconnects
    High Availability
    AMD
    Intel

    News
    Submit News/Article/PR
    Latest News
    Newsletter
    News Archives
    Search Archives
    Reference
    Featured Articles
    Beginners
    Whitepapers
    Documentation
    Software
    Lists/Newsgroups
    Books
    User Groups
    Higher Education
    Cluster List
    Linux HPC News Update
    Stay current on Linux related HPC news, events and information.
    LinuxHPC Newsletter

    Other Mailing Lists:
    Linux High Availability
    Beowulf Mailing List
    Gelato.org (Linux Itanium)

    Linux HPC Links
    Beowulf.org
    Beowulf Users Group
    Cluster Monkey
    High Performance Computing Clusters
    Gelato.org
    The Aggregate
    Top500
    Cluster Benchmarks
    Cluster Computing Info Centre
    Coyote Gultch
    Linux Clustering Info Ctr.
    Robert Brown's Beowulf Page
    Sourceforge Cluster Foundry
    HPC DevChannel
    OpenSSI
    Grid-Scape.org
    HPCWire
    SuperComputingOnline
    HPC User Forum
    Scalability.org
    Gridtech
    News Feed
    LinuxHPC.org.org has an RSS/RDF feed if you wish to include it on your website.
    LinuxHPC.org
    Home
    About
    Contact
    Mobile Edition
    Sponsorship

    Linux Cluster RFQ Form
    Reach Multiple Vendors With One Linux Cluster RFQ Form. Save time and effort, let LinuxHPC.org do all the leg work for you free of charge. Request A Quote...

    LinuxHPC.org is Looking for Interns - If you have experience with Linux clusters and/or cluster applications and you're interested in helping out with LinuxHPC.org let me know. I can promise the experience will be rewarding and educational, plus it'll make good resume fodder. You'll work with vendors and cluster users from around the world. We're also looking for writers and people to do hardware/software reviews. Contact Ken Farmer.

    Latest News

    IDC Reports Latest Supercomputer Ranking
    Wednesday September 10 2003 @ 12:19PM EDT

    FRAMINGHAM, Mass. -- IDC today reported the newest performance ratings of supercomputing and high-performance computing systems based on its enhanced metrics for the IDC Balanced Rating. IDC also predicted that the global high-performance computing market would average 6.1 percent annual growth to reach $6.3 billion in 2007.

    Supercomputers, or High-Performance Computers (HPC), are systems designed or configured to support scientific research, engineering design, and other computationally intense applications.

    In the rankings of 1,599 high-performance computers, Japan's 'Earth Simulator,' incorporating technology from NEC Corp., retained its lead as the top-ranked system based on the new rankings, which combine processor performance, memory performance, and system scalability measures. Multiple vendors exhibited performance leadership in specific segments of the market, according to Earl Joseph II, IDC Research Vice President, Worldwide Systems and Servers.

    · In the Capability class of HPC systems, computers used to tackle the most demanding individual problems (the traditional "supercomputer" segment of the market), IBM had the largest representation of the leading 200 computers (36 percent of the systems), followed by SGI (19 percent). Cray showed renewed momentum and had the highest-ranked computer in the Single Computer subcategory, while NEC led in the Cluster subcategory. Relative newcomer Linux Networx showed impressive momentum in the Capability class.

    · In the Enterprise class of capacity systems, priced at $1 million or more and used primarily for running many small to medium size jobs, Sun Microsystems led with 59 percent of the 200 highest-ranked systems, followed by Hewlett-Packard with 22 percent. IBM had the best-ranked computer in this segment.

    · In the Divisional capacity class of systems selling for $250,000 to $999,999, Hewlett-Packard led with 48 percent of the highest-ranked 200 systems, and SGI was second with 34 percent of the systems. SGI had the highest-ranked computer in this segment.

    · Among Departmental capacity class computers selling below $250,000, Hewlett-Packard was the leader with 53 percent of the highest-ranked 200 systems. SGI had the highest-ranked computer in this segment.

    "High performance computers are vital for national defense, industrial competitiveness and scientific progress, but in recent years users have pressed for better tests to help guide purchasing decisions," Joseph said. "The IDC Balanced Rating, developed with extensive input from the HPC community, includes multiple standard industry metrics to provide a more balanced view than rankings based on any single number." Among the metrics included are: Linpack (Rmax); SPECfp_Rate (base 2000); SPECint_Rate (base 2000); and STREAM TRIAD. Joseph expects additional tests to be added over time.

    "On behalf of IDC, I want to thank the HPC community and the SPEC committee for their excellent ideas for improving our rating methodology," he said. Joseph invited additional suggestions to be forwarded to him at hpc@idc.com.

    "User organizations need a source of common system specification data and top-level tools for sorting through various aspects of the technologies to better understand and explain how different computer architectures match their requirements. The IDC Balanced Rating provides one such top-level analysis and ranking based on processor, memory and scaling characteristics," added Christopher Willard, IDC Research Vice President, Worldwide Systems and Servers.

    The IDC Balanced Rating focuses on three broad performance areas:

    · Processor performance. The desired measurement is the speed at which the processors could generate results if they were kept fully occupied with work. For the IDC Balanced Rating, SPECfp_rate_base2000, SPECint_rate_base2000 and Linpack system Rmax results are used. In the future additional processor metrics will be incorporated.

    · Memory system capability. A computer's actual performance is dependent on how much data can be moved into and out of the processors in a given time period. For the IDC Balanced Rating, main memory peak bandwidth and the STREAM TRIAD benchmarks are used.

    · Scaling capability. A number of factors relate to how a large system can be configured, and how performance is affected by the design. For the IDC Balanced Rating, total system memory and interconnect bandwidths are used. In the future, latency and other metrics will be integrated into the benchmark based on the availability of consistent data on HPC computers.

    "Predicting the performance characteristics of a computer for your specific application codes is a very complex process and buyers often narrow the field of choices by first deciding on the basic computer architecture type, e.g., a single computer vs. a cluster, large nodes vs. small nodes, RISC vs. vectors, etc. Any broad ranking list like this one is often more useful in the acquisition of lower priced computers since you can't afford to spent a million dollars to evaluate buying a $200,000 system," said Willard.

    The full list of 1,599 computers is available without charge at: www.idc.com/hpc. The same website includes a tool that allows site visitors to custom-rank the listed HPC systems according to their own computing priorities.

    < Dartmouth: An Open-Source Virtual Secure Coprocessor based on TCPA | Streamline Computing Announces Availability of its DDT Parallel Debugger for the SGI Altix >

    Sponsors




    WinHPC.org


    Stay up-to-date on the Progress of Microsoft's Windows Compute Cluster Server 2003

    Golden Eggs
    (HP Visual Diagram and Config Guides)
    Integrity:
    HP Integrity Family Portrait, IA64
    rx1620 1600MHz 2P MSA1000 Cluster IA64
    rx2620 1600MHz 2P MSA1000 Cluster IA64
    rx4640 1600MHz 4P MSA1000 Cluster IA64
    ProLiant:
    DL140 3060MHz 2P IA32
    DL140 G2 3600MHz 2P EM64T
    DL145 2600MHz 2P Opteron
    DL145 G2 2600MHz 2P Opteron Dual Core
    DL360 G4 3400MHz 2P EM64T
    DL360 G4p 3800MHz 2P EM64T
    DL380 G4 3800MHz 2P EM64T
    DL385 2800MHz 2P Opteron Dual Core
    DL560 3000MHz 4P IA32
    DL580 G3 3330MHz 4P EM64T
    DL585 2800MHz 4P Opteron Dual Core
    Clusters:
    CP3000 32x DL140G2 & DL360G4p GigE EM64T
    CP4000 32x DL145G2 GigE Opteron, Dual Core
    CP4000 64x DL145 GigE Opteron
    CP4000 102x DL145 GigE Opteron
    CP4000 32x DL145 Myri Opteron
    Rocks Cluster 16-22 DL145 Opteron
    Rocks Cluster 30-46 DL145 Opteron
    Rocks Cluster 64-84 DL145 Opteron
    LC3000 GigaE 24-36 DL145 Opteron
    LC3000 Myri 16-32x DL145 Opteron
    LC3000 GigaE 16-22x DL145 Opteron
    LC2000 GigaE 16-22x DL360G3 Xeon
    Storage:
    MSA500 G2, SCSI
    MSA1500 24TB, SCSI and SATA
    Misc:
    Dual Core AMD64 and EM64T systems with MSA1500









    Linux Magazine
    At Newstands Now!Linux Magazine

    Click for King, North Carolina Forecast





       


    Home About News Archives Contribute News, Articles, Press Releases Mobile Edition Contact Sponsorship Search Privacy
         Copyright © 2001-2005 LinuxHPC.org
    Linux is a trademark of Linus Torvalds
    All other trademarks are those of their owners.
        
      SpyderByte.com ;Technical Portals