Reprints from my posting to SAN-Tech Mailing List and ...

2011/06/23

[san-tech][03204] Graph 500 Complete Results - June 2011公開

Date: Thu, 23 Jun 2011 18:04:06 +0900
------------------------------------------------
Graph 500 Complete Results - June 2011が公開されました

Complete Results - June 2011, Graph 500
  http://www.graph500.org/june2011.html

エントリー数が、前回 (Nov. 2010) の 9から 28に大幅に増えています。

コア数よりも、インターコネクトやノード数の詳細を記して欲しいです。
特に通信ライブラリーの情報は必須ですね。
※う~、nodes/processors/coresでの表記が混在しているのは良くないです。

せめて Top 500レベルの情報が欲しいです。

------------------------------------------------
2011/07/26

"Lomonosov Supercomputer Sets World Record for Performance on Graph500 Benchmark"
 30 June 2011
  http://www.t-platforms.com/about-company/press-releases/186-graph500.html
.....
  "... When the test results were analyzed, it was found that the Lomonosov,
   a Russian supercomputer designed by T-Platforms, had set a new performance
   record, reaching a level of 43.5 GE/s (billion edges processed per second)."

"Interview: Russian Supercomputer Outperforms All Others on Graph500"
 07.15.2011, insideHPC
  http://insidehpc.com/2011/07/15/interview-russian-supercomputer-outperforms-all-others-on-graph5000/

  "This week T-Platforms announced that Russia's most powerful supercomputer
   at Moscow State University outperformed all competitors during the recent
   Graph500 benchmark tests. To learn more, I caught up with Anton Korzh,
   a systems architect at T-Platforms."

MSU SUPERCOMPUTERS: "LOMONOSOV"
  http://hpc.msu.ru/?q=node/59
Moscow University Supercomputing Center
  http://hpc.msu.ru/

T-Platforms
  http://www.t-platforms.com/


Complete Results - June 2011
  http://www.graph500.org/june2011.html

Rank: Machine, Owner,
   Problem Size, TEPS (Traversed Edges Per Second)
1: Intrepid (IBM Blue Gene/P, 32,768 nodes / 131,072 cores),ANL,
   38, 18,508,000,000
2: Jugene (IBM Blue Gene/P, 32,768 nodes / 131,072 cores),
   Forschungszentrum Julich,
   38, 18,416,700,000
3: Lomonosov (MPP, 4096 nodes / 8192 cores), Moscow State University,
   37, 43,471,500,000
4: Hopper (Cray XE6, 1800 nodes / 43,200 cores),LBL,
   37, 25,075,200,000
5: Franklin (Cray XT4, 4000 nodes / 16,000 cores), LBL,
   36, 19,955,100,000
================================================
[san-tech][03190] Slidecast: "Convey Hybrid Core Computing Rocks the Graph 500", 06.18.2011, insideHPC
Graph 500一般に関しては
[san-tech][02900] Cray XMT (Massively Multithreaded Platform) プレスリリース等
> "まったく異なるスパコンの性能評価「Graph500」"
>  2011/01/12
>  SC10 - HPC Challenge、Green500とGraph500 (3)
>   http://journal.mycom.co.jp/articles/2011/01/12/sc10_hpc_challenge/002.html
※これも安藤さんの解説記事です。
[san-tech][03197] Graph 500ベンチマークをソフトウェアで高速化, E.T. International, Inc. (ETI)
[san-tech][03199] Re: Graph 500ベンチマークをソフトウェアで高速化, E.T. International, Inc. (ETI)

0 件のコメント:

コメントを投稿