Physics Analysis Centres

1
Physics Analysis Centres

• Pierre Savard
• University of Toronto and TRIUMF
• Toronto
• July 2004

2
Physics Analysis Centre

• Two definitions
• A place where a significant group of people get
  together to work on physics analyses
• A computing centre where significant resources
  are dedicated to high-level physics analyses
• The two are factorizable

3
Motivation

• I asked myself two questions
• What physics do I want to do?
• I want to discover the Higgs! (VBF)
• I want to find physics beyond the Standard Model
  (Jets Etmiss)
• How will I do physics analysis in ATLAS
• Very high-profile analysis implies working in a
  very competitive environment
• Can I do this without being based at CERN?
• What computing resources will I have at my
  disposal to compete with the rest of the
  collaboration?

4
Recent Experience from CDF

• Top cross section in dilepton channel
• two teams of a dozen physicists each
• one team started more than a year after the other
• analyses required very large computing resources
• Very large MC samples (10 million Z, gt1 Million
  dibosons, gt1 million top, etc
• Had to spin over many datasets (all jet samples,
  photons, inclusive leptons)
• A lot of grinding with high-level analysis code
• Required frequent trips to fermilab
• First Tevatron top physics publication

5
A Canadian Higgs Search Group?

• Can we assemble a Canadian group with the
  critical mass necessary to play a leading role in
  a key search channel?
• Make a clearly identifiable Canadian contribution
  to Higgs physics/discovery
• Group would have a contingent at CERN but with a
  critical mass in Canada
• Would build on significant experience and
  knowledge of calorimetry and forward jet tagging

Can do the same with a high-profile search
channel e.g. Jets Etmiss
6
Physics Analysis Centre (I)

• If we have the critical mass for one or two large
  analysis groups, we could focus the effort in one
  Canadian location with
• dedicated offices with terminals/printer etc.
• 1 or more meeting rooms with video, projector
  etc.
• Would also be useful to Canadians working
  on other topics
• Allows fermilab model of frequent
  relatively short/inexpensive trips. Students not
  too far from advisors

7
Physics Analysis Centre (II)

• A few notes on CDF computing resources
• Have 1 dedicated processing farm (will have two)
• Have one dedicated simulation farm
  (Toronto/Alberta)
• Have central analysis facility (more than twice
  as large as reconstruction farm). We have a few
  large offsite analysis facilities. Analysis
  facilities are available to all CDF
  collaborators.
• Analysis CPU/Disk requirements are significant
• Have we assigned a very significant fraction of
  dedicated disk and cpu resources for Canadian
  analyses?
• If not, should we plan on requesting addtional
  resources?

8
My Conclusions (from previous TRIUMF talks)

• ATLAS a key component of Canadian subatomic
  physics program
• Will answer some of the fields great questions
• Origin of mass
• Is there new physics at the TeV energy scale
• Canadian and TRIUMF physicists
• have already made significant contributions to
  ATLAS physics studies
• will play an essential role in commissioning and
  understanding of ATLAS detector
• will play important role in early, high profile
  physics analyses

9
My Conclusions (for today)

• To play a leading role as a Canadian group in
  some high profile analyses (search for Higgs,
  physics beyond SM), well need to build critical
  mass here in Canada
• A centralized Canadian location would help these
  efforts
• We will need dedicated computing resources to
  compete with other groups
• Will we have them? Should we think of making a
  specific request for additional resources?