5
Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN Deputy Director, Research Computing Center, Moscow State University Corresponding member of Russian Academy of Sciences, [email protected] 25 March 2010, Brussels

Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN

  • Upload
    talia

  • View
    30

  • Download
    1

Embed Size (px)

DESCRIPTION

Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN Deputy Director, Research Computing Center, Moscow State University Corresponding member of Russian Academy of Sciences, [email protected] 25 March 2010, Brussels. - PowerPoint PPT Presentation

Citation preview

Page 1: Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN

WorkshopEU – Russia Joint Call in High Performance Computing

Prof. VLADIMIR VOEVODIN

Deputy Director, Research Computing Center, Moscow State University

Corresponding member of Russian Academy of Sciences,

[email protected]

25 March 2010, Brussels

WorkshopEU – Russia Joint Call in High Performance Computing

Prof. VLADIMIR VOEVODIN

Deputy Director, Research Computing Center, Moscow State University

Corresponding member of Russian Academy of Sciences,

[email protected]

25 March 2010, Brussels

Page 2: Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN

Moscow State UniversityMoscow State University1755 – 20101755 – 2010

30+ Faculties30+ Faculties350+ Departments350+ Departments

5 major Research Institutes5 major Research Institutes

More than 40 000 students, More than 40 000 students, 2500 full doctors, 6000 PhDs,2500 full doctors, 6000 PhDs,

1000+ full professors,1000+ full professors,5000 researchers.5000 researchers.

Moscow State UniversityMoscow State University1755 – 20101755 – 2010

30+ Faculties30+ Faculties350+ Departments350+ Departments

5 major Research Institutes5 major Research Institutes

More than 40 000 students, More than 40 000 students, 2500 full doctors, 6000 PhDs,2500 full doctors, 6000 PhDs,

1000+ full professors,1000+ full professors,5000 researchers.5000 researchers.

Page 3: Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN

There are 25 Doctors of Sciences in RCCThere are 25 Doctors of Sciences in RCCThere are 25 Doctors of Sciences in RCCThere are 25 Doctors of Sciences in RCC

Research Computing Center, MSUResearch Computing Center, MSU1955 – 20101955 – 2010

20 Laboratories,20 Laboratories,220+ Researchers,220+ Researchers,

50 PhDs and 25 Full Doctors,50 PhDs and 25 Full Doctors,

RCC MSU – Supercomputing Center #1 in RussiaRCC MSU – Supercomputing Center #1 in Russia

Research Computing Center, MSUResearch Computing Center, MSU1955 – 20101955 – 2010

20 Laboratories,20 Laboratories,220+ Researchers,220+ Researchers,

50 PhDs and 25 Full Doctors,50 PhDs and 25 Full Doctors,

RCC MSU – Supercomputing Center #1 in RussiaRCC MSU – Supercomputing Center #1 in Russia

Page 4: Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN

Performance analysis tools for HPCPerformance analysis tools for HPC(what problems should the project address?)(what problems should the project address?)

• Most supercomputers have extremely low efficiency for a very wide range of applications.

• Most users have no information about their programs after submitting to a queue…

• “Low efficiency” can’t be explained by one reason, this is a complex problem…

• Sources of losses: policies and quotas of batch systems, RTSs, compilers, communications overheads, load imbalance, Amdahl’s law, a memory wall…

• There is neither a unified approach nor a software tool to detect sources of losses for users (for a particular task) and system administrators (for the whole supercomputer).

Page 5: Workshop EU – Russia Joint Call in High Performance Computing Prof. VLADIMIR VOEVODIN

Performance analysis tools for HPCPerformance analysis tools for HPC(some key points of the project)(some key points of the project)

• Create an integrated environment that combines existing and new tools from the level of batch systems down to hardware monitors.

• Detect all sources of losses on the level of a particular task, a particular user, and the entire supercomputer.

• Collect, analyze, filter and display data in a scalable way up to exascale range systems.

• Analyze instrumented as well as non-instrumented tasks.

• Target architecture – clusters of SMP nodes with multicore processors and GPUs.