Please read these instructions before posting any event on Fermilab Indico

Indico will be unavailable on Wed., Dec 18th from 7 - 7:30am due to server maintenance.

Snowmass CompF4 Topical Group Workshop

US/Pacific
Frank Wuerthwein (UCSD), Meifeng Lin (BNL), Wahid Bhimji (NERSC, Berkeley Lab)
Description

Meeting to discuss the Snowmass CompF4 Topical Group Report. 

Register in advance for this meeting.

After registering, you will receive a confirmation email containing information about joining the meeting.

  • Thursday, 7 April
    • 09:00 09:20
      Intro and topical group overview 20m
      Speaker: Wahid Bhimji (NERSC, Berkeley Lab)
    • 09:20 09:50
      Networking
      Conveners: Chin Guok (ESnet), Eli Dart (ESnet / LBNL), Shawn McKee (Univ. of Michigan)
    • 09:50 10:20
      Edge Services
      Convener: Ofer Rind (BNL)
    • 10:20 10:50
      Break 30m
    • 10:50 11:20
      Analysis Facilities: Discussion: Analysis Facilities
      Conveners: Gordon Watts (University of Washington), Mark Neubauer (University of Illinois at Urbana-Champaign), Oksana Shadura (University of Nebraska Lincoln (US))
      • 10:50
        Analysis Cyberinfrastructure: Challenges and Opportunities 7m
        Speaker: Kevin Lannon (University of Notre Dame)
      • 10:57
        Analysis Facilities for HL-LHC (DOE) 7m
        Speaker: Doug Benjamin (Brookhaven National Laboratory (US))
      • 11:04
        Analysis Facilities for HL-LHC (NSF) 7m
        Speakers: Lincoln Bryant (University of Chicago), Robert Gardner (University of Chicago)
      • 11:11
        Leveraging Cloud Native for designing, building and operating analysis facilities for USCMS 7m
        Speaker: Maria Acosta
    • 11:20 11:50
      AI Hardware
      Conveners: Javier Duarte (University of California San Diego), Nhan Tran (FNAL)
    • 11:50 12:20
      Break 30m
    • 12:20 12:50
      Processing
      Conveners: Ian Fisk (Flatiron Institute), Meifeng Lin (BNL)
      • 12:20
        Processing Overview 5m
      • 12:25
        Evolution of resource provision models and the role of HPC 10m
        Speaker: Wahid Bhimji (NERSC, Berkeley Lab)
      • 12:35
        Heterogeneous hardware adoption and portability 5m
        Speaker: Matti Kortelainen (Fermilab)
      • 12:40
        Authorization and Authentication Evolution 5m
        Speaker: Brian Bockelman (Morgridge Institute for Research)
      • 12:45
        Discussion 5m
    • 12:50 13:20
      Storage
      Conveners: Bo Jayatilaka (Fermilab), Carlos Maltzahn (UC Santa Cruz), Peter van Gemmeren (ANL)
    • 13:20 13:50
      Discussion 30m
    • 13:50 14:00
      Plan for tomorrow 10m
    • 09:00 10:00
      AI Hardware
    • 09:00 10:00
    • 09:00 10:00
      Edge Services
    • 09:00 10:00
      Networking: Networking Discussion
    • 09:00 10:00
      Processing
    • 09:00 10:00
      Storage: Topic Discussion
      Conveners: Bo Jayatilaka (Fermilab), Carlos Maltzahn, Peter van Gemmeren (ANL)
      • 09:00
        Upcoming Storage Features in ROOT 15m
        Speakers: Jakob Blomer (CERN), Philippe Canal (FERMILAB)
      • 09:15
        Computational Storage 15m

        What is Computational Storage? Who is Samsung's Memory Solutions Lab, what have we done, and what do we think we can do in the upcoming years.

        Speaker: Vitorio Cargnini (Samsung)
      • 09:30
        Distributed data management at JGI 15m

        The DOE Joint Genome Institute is a user facility that generates genome sequence data and analysis products for the environmental research community. I’m an effort to be more resilient to outages, JGI has recently shifted to a distributed computing model. I will briefly describe our hierarchical data management system that complements our centralized workflow system.

        Speaker: Kjiersten Fagnan (LBL)
      • 09:45
        Data Context and Sharded Data - Erasure Coding for HDD Computational Storage 15m

        Discussion of the need for data context at the storage device to perform data layout aware computations on the device. An example of structured data stored via an erasure coded data availability scheme and how it destroys data context at the device. Possible solutions are described.

        Speaker: Philip Kufeldt (Seagate)
    • 10:00 10:20
      Break 20m
    • 10:30 11:45
      Bulleted list of priorities from each topic: Analysis Facilities; Storage: Processing: (20-25 mins each)

      Things agreed on - structure of report
      Things in limbo - for further workshops

      Convener: Wahid Bhimji (NERSC, Berkeley Lab)
    • 11:40 12:20
      Break 40m
    • 12:20 13:35
      Bulleted list of priorities from each topic: AI Hardware; Edge Services; Networking (20-25 mins each)

      Things agreed on - structure of report
      Things in limbo - for further workshops

      Convener: Wahid Bhimji (NERSC, Berkeley Lab)
    • 13:40 14:00
      Wrap up and Next Steps Towards Report 20m