Benchmark WG Meeting

Wednesday 27 Feb 2008, 17:00 → 18:00 Europe/Zurich

1/1-025 (CERN)

Helge Meinhard (CERN-IT)

Access code 2008-02-27-participant.txt

CPU Benchmark WG Meeting Notes, February 27, 2008


Present:
Alex Iribarren, Helge Meinhard, CERN/IT
Franco Brasolin, CERN/ATLAS
Hubert Degaudenzi, CERN/LHCb
Peter Hristov CERN/ALICE,
Michele Michelotto, INFN Padova
Gabriele Benelli CERN/CMS
Martin Bly, RAL
Sandy Philpott, JLab
Ian Gable, U Victoria

1. Comments on minutes from Jan 23

• no comments

2. Review of action items

• Gabrielle had been in contact with Peter Wegner to test his Happertown Machine. Peter was at CERN thus making the process easy. Peter  was unable to make the meeting due to travel.
  
• The main item was more work by the experiment contacts (more in the next section).

3. Progress from experiments.

Gabriele Benelli (CMS)

• There has been some flux in the way settings are being distributed within CMS right now. Gabriele expects this to stabilize shortly.
• Has been running across lxbench01 to 07 using different types of process for each core. He is also studying the effects of having N processes running on a M core machine where N is {1,...,M}.
• Helge suggest that Gabrielle first concentrate on running the same process on all available cores as it is the most pressing issue.
• Gabrielle will package the code he is using so that it can be pulled out for use later.

Hubert Degaudenzi (LHCb)

• Has sorted out some software problems and is ready to run on one of the lxbench machines.

Peter Histov (Alice)

• Ready to run in 64 bit mode.
• He was experiencing some problems with using too much memory (6GB per core) now cut down to less then 2GB
• Has some preliminary results, not ready to show yet as he hasn't done a full run.

Franco Brasolin (ATLAS)

• Results already posted to mailing list.
  

4. Action Item

• Ian/Manfred to start organizing results on the wiki.
• CMS to have results by next meeting.

5. Next Meeting

• The next meeting will be held March 12th 16:00 UTC.