徒然なるままに

Mail: topography "AT" mail.goo.ne.jp

High Productivity Computing Systems(HPCS) III: IBM

2006-01-08 | SuperComputer
High Productivity Computing Systems(HPCS) VI: IBM 2, 2006-04-07
==========
[06/03/12]
"IBM and Agilent Technologies Launch $30 Million Interconnect Program", 11 Sep 2003
 http://www.ibm.com/press/us/en/pressrelease/5828.wss
 "IBM and Agilent Technologies Inc. (NYSE: A) today announced that they are teaming up to advance
  the performance of optical interconnects as part of a multi-year, $30 million project with
  the Defense Advanced Research Projects Agency (DARPA)."
 "Key goals of IBM and Agilent's program include:

   aggregate data rates up to 1 Tbps per module;
   high per-channel rates for optimal cost and power consumption --
    up to 15 Gbps for parallel links, and up to 40 Gbps for components;
   small form factor modules less than two square centimeters in area;
   ultra low power (5 to 10 milliwatts/Gbps); and
   high levels of integration to allow tight coupling with computer chips for optical transceivers and waveguide interconnects.

 "The team is expected to demonstrate the enormous communication bandwidths needed for
  future servers such as those under the DARPA High Productivity Computing Systems (HPCS) initiative.
  IBM's HPCS project (called PERCS, or "Productive, Easy-to-use, Reliable Computing System") will
  investigate the use of hundreds or thousands of terabit-class optical interconnects to overcome
  the bandwidth limitation of today's server interconnects."

"Exploitation of optical interconnects in future server architectures",
 Alan F. Benner, IBM Systems and Technology Group, et al.,
 IBM Journal of Research and Development, Volume 49, Number 4/5, 2005.
 http://www.research.ibm.com/journal/rd/494/tocpdf.html
==========
[06/01/12]
 IBM Petascale Strategy Workshop, May 3-4, 2005
  IBM Research Advanced Computing Technology Center
  http://researchweb.watson.ibm.com/actc/workshop.shtml
==========
[06/01/08]
DARPA(Defense Advanced Research Projects Agency:米国防総省高等研究計画局)
High Productivity Computing Systems (HPCS)での IBMの動向です:

IBMのプレスリリース:
"IBM wins DARPA funding", 10 Jul 2003
 http://www-03.ibm.com/press/us/en/pressrelease/5487.wss

"米国防総省のスパコン開発プログラム第2段階に突入,
  米IBM,米Cray,米Sunに総額1億4600万ドルの奨励金を授与", 2003年07月11日
 http://itpro.nikkeibp.co.jp/free/ITPro/USNEWS/20030710/9/
  "・IBM社
    多数の異なる作業に適用可能なコンピュータを設計するためのPERCS
   (Productive,Easy-to-use,Reliable Computing Systems)プログラムに
   向けた研究開発を行なう。5330万ドルが支給された。"

IBMのリーダは IBM Austin Research Laboratoryの Mootaz Elnozahy博士です。

PERCS(Productive Easy-to-use Reliable Computer Systems)の Webページは IBMに
ないようですが、開発中のプログラミング言語 X10については:
The X10 Programming Language
 http://domino.research.ibm.com/comm/research_projects.nsf/pages/x10.index.html
PERCSについてはピッツバーグ大学に:
 University of Pittsburgh P.E.R.C.S. Project
  http://www.cs.pitt.edu/percs/
  Research
   Compiled Communication for Parallel Programs
   Dual Electronic-Optical Interconnection Networks
   Optical Circuit Switching Networks
もあります。

国際会議等での発表:
 HPCS SC2005 Workshop, November 14, 2005
  http://highproductivity.org/Nov2005.htm
  "Toward Economically Viable Peta-Scale Computing"
    Mootaz Elnozahy, IBM Austin Research Lab
    http://highproductivity.org/Nov05/2-IBM%20Talk.pdf
 Conference on Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA), October 2005
  "X10: An Object-Oriented Approach to Non-Uniform Cluster Computing"
   Philippe Charles, Christopher Donawa, Kemal Ebcioglu, Christian Grothoff(UCLA), Allan Kielstra, Christoph von Praun Vijay Sarsawat and Vivek Sarkar
   http://www.research.ibm.com/people/p/praun/oopsla_onward05.pdf
   http://portal.acm.org/citation.cfm?id=1094811.1094852&coll=ACM&dl=ACM&type=series&idx=1094811&part=Proceedings&WantType=Proceedings&title=Conference%20on%20Object%20Oriented%20Programming%20Systems%20Languages%20and%20Applications&CFID=61763287&CFTOKEN=64823959
 Workshop on Application Specific Processors, September 22, 2005
  Keynote Speech:
  "IBM PERCS Project: Hardware-software co-design of a future supercomputer
   for high programmer productivity"
    Kemal Ebcioglu, IBM Research
    http://dogbert.eng.umd.edu/wasp05/keynote.pdf
 AHPCRC - DARPA Programming Models Conference, 12-14 September 2005
  http://www.ahpcrc.org/conferences/PGAS/index.html
  "An Object-Oriented Approach to PGAS Programming"
    Vivek Sarkar, IBM T.J. Watson Research Center
    http://www.ahpcrc.org/conferences/PGAS/presentations/Sarkar_X10.pdf
 PMUA 2005: Workshop in programming models for HPCS ultra-scale applications, June 21, 2005
  http://crd.lbl.gov/~oliker/pmuahome.html
   "X10 --- a New Programming Model for Productive Scalable Parallel Programming"
    Vivek Sarkar, IBM
    http://crd.lbl.gov/%7eoliker/pmua/IBM-X10-Sarkar.pdf
 Language Runtimes '04, October 28, 2004
  http://www.aurorasoft.net/workshops/lar04/
  Presentations:
   http://www.aurorasoft.net/workshops/lar04/Presentations/LaR04Presentations.htm
  "X10: A Programming Model for Hierarchical Heterogeneous Parallelism"
    Vivek Sarkar, IBM T.J. Watson Research Center
    http://www.aurorasoft.net/workshops/lar04/Author_Files/Presentations/Vivek_Sarkar_LaR_04_V2.pdf
    http://www.aurorasoft.net/workshops/lar04/Author_Files/Papers/Vivek_Sarkar_LaR_04_Paper_V1.pdf
 Invitational Workshop on the Future of Virtual Execution Environments, September 15-17, 2004
  http://www.research.ibm.com/vee04/talks.html
  "Language and Virtual Machine Challenges for Large-Scale Parallel Systems"
    Vivek Sarkar, IBM Research
    http://www.research.ibm.com/vee04/Sarkar.pdf
    ☆この資料によると、PERCSの System Software(Kernel)は K42だそうです:
     The K42 Project
      http://researchweb.watson.ibm.com/K42/
 The Coalition for Academic Scientific Computation Meeting
  "PERCS: IBM Effort in HPCS", Augst 06, 2003
    Mootaz Elnozahy, IBM Austin Research Lab
    http://www.ncsc.org/casc/meetings/Vision-Public.pdf


関連 Web記事:
"IBMがスーパーコンピュータに賭ける理由", 2004/01/08
 http://japan.cnet.com/interview/story/0,2000050154,20063533,00.htm
  "TRIPS (Tera-op Reliable Intelligently Adaptive Processing System)では、
   IBMとテキサス大学が、毎秒1兆のアプリケーションを動かせる
  「チップ上のスーパーコンピュータ」を研究中だ。"
"米IBM、ワンチップにスーパーコンピュータの性能を詰め込む", 2003/08/27
 http://www.japan.cnet.com/news/ent/story/0,2000047623,20060627,00.htm
 TRIPS (The Tera-op, Reliable, Intelligently adaptive Processing System)
 The University of Texas at Austin, Department of Computer Sciences
  http://www.cs.utexas.edu/users/cart/trips/
  TRIPSのメンバーが所属する:
  CART (Computer Architecture and Technology Laboratory)
   http://www.cs.utexas.edu/users/cart/


関連エントリー:
 High Productivity Computing Systems(HPCS) I: 概略, 2006-01-08


最新の画像もっと見る

コメントを投稿

ブログ作成者から承認されるまでコメントは反映されません。