Choose timezone
Your profile timezone:
Google/ATLAS/CERN openlab tutorial on advanced performance monitoring
→
Europe/Zurich
CERN
CERN
Description
Google/ATLAS/CERN openlab tutorial on advanced performance monitoring
October 17-18 2012, CERN
Registration: https://edh.cern.ch/Document/Personnel/TRN/new?course=123GPM01
https://vidyoportal.cern.ch/flex.html?roomdirect.html&key=QBbcqazWkq1Y
Profiting from David Levinthal’s* visit at CERN, a tutorial on advanced performance monitoring is being organized jointly by Google, ATLAS and CERN openlab. During the two day course the participants will be able to get an advanced understanding of performance tuning theory and practice on modern PCs. In particular, David will demonstrate a novel tool called GOODA, the Generic Optimization Data Analyzer. GOODA leverages the widely available perf subsystem in Linux and is co-developed by Google and the ATLAS collaboration. The hands-on labs will consist in part of pre-developed exercises and in part of an analysis of the attendees’ own applications. These examples will be used to explain how to debug and fix real performance issues.
The tentative agenda is as follows:
*) Dr. David Levinthal is one of the world’s top experts in the PC performance monitoring field and a Staff engineer at Google. For a large part of his career he has worked as a Principal Engineer at Intel, designing and developing performance tuning software and hardware. Earlier, David worked at CERN where he was a physicist in the ALEPH and CCOR experiments. Having followed that link many years later, David has a history of collaboration with openlab on performance tuning activities. David’s main occupation today is optimization work for Google which includes driving the development of GOODA – the Generic Optimization Data Analyzer.
October 17-18 2012, CERN
Registration: https://edh.cern.ch/Document/Personnel/TRN/new?course=123GPM01
https://vidyoportal.cern.ch/flex.html?roomdirect.html&key=QBbcqazWkq1Y
Profiting from David Levinthal’s* visit at CERN, a tutorial on advanced performance monitoring is being organized jointly by Google, ATLAS and CERN openlab. During the two day course the participants will be able to get an advanced understanding of performance tuning theory and practice on modern PCs. In particular, David will demonstrate a novel tool called GOODA, the Generic Optimization Data Analyzer. GOODA leverages the widely available perf subsystem in Linux and is co-developed by Google and the ATLAS collaboration. The hands-on labs will consist in part of pre-developed exercises and in part of an analysis of the attendees’ own applications. These examples will be used to explain how to debug and fix real performance issues.
The tentative agenda is as follows:
- Day 1 morning – Theory and lectures
- Day 1 afternoon – GOODA exercises
- Homework: profile your own app
- Day 2 – Discussion of collected profiles and solving real performance issues
*) Dr. David Levinthal is one of the world’s top experts in the PC performance monitoring field and a Staff engineer at Google. For a large part of his career he has worked as a Principal Engineer at Intel, designing and developing performance tuning software and hardware. Earlier, David worked at CERN where he was a physicist in the ALEPH and CCOR experiments. Having followed that link many years later, David has a history of collaboration with openlab on performance tuning activities. David’s main occupation today is optimization work for Google which includes driving the development of GOODA – the Generic Optimization Data Analyzer.
The agenda of this meeting is empty
