WP12 General Meeting

→ Europe/Zurich
Zoomiverse

Zoomiverse

Description
Join Zoom Meeting
https://desy.zoom.us/j/83910221339

Meeting ID: 839 1022 1339
Passcode: 089063

WP12 Meeting Minutes, 2023-03-15

Present: Graeme Stewart, Frank Gaede, Juan Carceller, Tao Lin, Anna Zaborowska, Anatolii Korol, Corentin Allaire, Peter McKeown, Hadrien Grasland, Swathi Sasikumar, Yizhou Zhang

Apologies: Andre Sailer, Thomas Madlener, John Back

Introduction

Post-Valencia general meeting proposed for 14 or 21 June.

ACTION: We will make a Google Doc where we can track who is coming to Valencia and also discuss the topics that people want to work on.

Turnkey Software

Skip presentation this time, with ideas for the hackathon on the agenda.

Corentin can't come to Valencia, unfortunately.

Simulation

ODD event (p7) is done through DDSim and displayed in Phoenix.

Track Reconstruction

SYCL vs. CUDA - syntax for writing code is similar. CUDA is better docucumented. Performance winner flip-flops - sometimes one ahead, sometimes another; not a clear winner.

EDM4hep integration with ACTS would be interesting

Particle Flow Reconstruction

Brief update on the slides, with ideas for hackathon - would want APRIL people to come.

AOB

See you all in Valencia!

There are minutes attached to this event. Show them.
    • 15:00 → 15:10
      Introduction and WP Status 10m
      Speakers: Frank-Dieter Gaede (Deutsches Elektronen-Synchrotron (DE)), Dr Graeme A Stewart (CERN)
    • 15:10 → 15:22
      Turnkey Software 12m
      Speaker: Thomas Madlener (Deutsches Elektronen-Synchrotron (DESY))
      • ML Shower FastSim integration into DD4hep / Geant4 and Key4hep. (DESY should be almost fully present for this, and AFAIU pretty much all pre-requisites are in place, including a prototype. Main points would be having this end-to-end and making sure Key4hep environment brings everything necessary. Potentially we can also try to figure out the basics of GPU support within Key4hep)

      • Gaussino integration (Depends a bit on how much time Graeme has, but in principle a lot of the smaller details could be handled here with a few hands on deck and someone who hands out directions)

      • Issue squashing and re-checking (Very much dependent again on the actual open issues then and on the people who are present, but quite a few of the Key4hep repos have rather old issues and cleaning them up would be something that can scale to quite a few people)

      • Better development workflows for Key4hep + documentation to ease on boarding of new people. (Very broad topic, can host people from all kinds of backgrounds and could in principle also be a solo project. I think here we could profit from some better documentation and also at least some sort of canonical approach)

      • ACTS integration in all its facets, e.g. EDM4hep interoperability, FW integration. (Only makes sense if someone from ACTS is actually present)

    • 15:22 → 15:34
      Simulation 12m
      Speaker: Anna Zaborowska (CERN)
    • 15:34 → 15:46
      Track Reconstruction 12m
      Speaker: Hadrien Benjamin Grasland (Université Paris-Saclay (FR))
    • 15:46 → 15:58
      Particle Flow Reconstruction 12m
      Speaker: John James Back (University of Warwick (GB))