Timetable: HEP-CCE: Scalable IO for Energy and Intensity Frontier Experiments

chaired by Philippe Canal (Fermilab), Peter Van Gemmeren (Argonne National Laboratory), Taylor Childers (Argonne National Laboratory), Samantha Tezak (Argonne National Lab)
from to (US/Central)
at Argonne - Bldg 240 ( 1404, 1405, 1406 & 1407 )
9700 Cass Ave, Lemont, IL 60439
Description High Energy Physics experiments continue to become more data and simulation intensive. There is an expected factor of ten gap (or more) between projected computing needs for HL-LHC experiments and projected growth of current HEP resources. In the US, High Performance Computing resources are going to be growing by more than an order of magnitude by 2021/2022 with the deployment of the US DOE's first exascale supercomputers. These resources are already becoming an important piece of the HEP computing landscape and will continue to become more important.

One challenge of using leadership computing resources is reading and writing data in scalable ways that does not lead to bottlenecks or performance penalties on the shared filesystems. This Workshop aims to bring together leading IO experts in the HEP field with experts from DOE ASCR Facilities to discuss how to move forward in the next years to make HEP software more friendly to millions of parallel threads accessing files on shared disks.
Material:
Go to day
  • Thursday, August 23, 2018
    • 09:00 - 10:30 Thursday Morning Intro
      Introduction to HPC IO & HEP IO
      Conveners: Dr. Taylor Childers (Argonne National Laboratory), Mr. Philippe Canal (FERMILAB)
      • 09:00 Brief Introduction to HPC IO 45'
        Speaker: Richard Zamora (Argonne)
        Material: Slides pdf file
      • 09:45 Intro to ROOT IO 45'
        Speaker: Mr. Philippe Canal (FERMILAB)
        Material: Slides pdf file
    • 10:30 - 11:00 Coffee Break
    • 11:00 - 14:00 Thursday Afternoon Intro
      Conveners: Dr. Taylor Childers (Argonne National Laboratory), Mr. Philippe Canal (FERMILAB)
      • 11:00 CMS IO Overview 40' ( )
        Speakers: Brian Bockelman, Brian Bockelman, Dr. Christopher Jones (Fermilab)
        Material: Slides pdf file
      • 11:40 ATLAS IO Overview 40' ( )
        Speaker: Dr. Peter van Gemmeren (ANL)
        Material: Slides powerpoint file pdf file
      • 12:20 NERSC Overview 30' ( )
        Speaker: Glen Lockwood (NERSC)
        Material: Slides pdf file
      • 12:50 OLCF Overview 30' ( )
        Speaker: Sarp Oral (Oak Ridge)
      • 13:20 ALCF Overview 30' ( )
        Speaker: Kevin Harms (Argonne)
        Material: Slides pdf file
    • 12:20 - 12:50 Working Lunch
    • 14:00 - 14:30 Coffee Break
    • 14:30 - 18:00 Thursday Evening Discussion
      • 14:30 Overview of HEP Activities in IO 30'
        Speaker: Jim Kowalkowski (Fermilab)
        Material: Slides powerpoint file pdf file
      • 15:00 HEP/HPC Tech Discussion 1h0'
        Discuss pros/cons of ROOT format vs HDF5 (others?). What's the future of Object Stores? Any IO tech in the pipeline to be aware of?
        Speakers: Dr. Taylor Childers (Argonne National Laboratory), Mr. Philippe Canal (FERMILAB)
    • 18:45 - 20:00 Dinner
      We have a reservation for 15 at The Craftsman in Naperville. Please fill out the doodle sent via email if you intend to attend.
  • Friday, August 24, 2018
    • 08:00 - 10:30 Friday Morning Discussion
      • 08:00 Object Stores 1h0'
        Quick introduction to Object Stores which the facilities indicate will be the leading scalable IO tool for the next ten years of Exascale.
        Speaker: Glen Lockwood
        Material: Slides pdf file
      • 09:00 HEP Transient Object Store 1h30'
        Deeper dive into R&D being done in HEP related to Object Stores. How can these solutions be generalized? How can LHC experiments utilized Object Stores to enable large scale IO on HPCs?
        Speaker: Rob Ross (Argonne)
    • 10:30 - 11:00 Coffee Break
    • 11:00 - 14:00 Friday Afternoon Discussion
      • 11:00 Compression 1h0'
        This session will be focused on how compression impacts scalable IO in HEP experiments. Most experiments store their data using compression. Can we utilize non-CPU resources to do this? What would the impact be on uncompressing on CPUs if compressed on GPUs? What tools can we use? ROOT? ZLIB? 
        Speakers: Dr. Peter van Gemmeren (ANL), Brian Bockelman, David Blythe (Argonne)
        Material: Slides link
      • 12:00 Working Lunch 30'
        Boxed lunches will be handed out during the last 30 minutes of Peter and Brian's discussion.
      • 12:00 Future Directions, Proposals, Action Items. 1h55'
        What action items have we identified. What joint projects can be tackled to prepare for Exascale use? ROOT IO? Object Stores?