Login

Heterogeneous Architectures in WLCG

3–5 Dec 2025
CERN
Europe/Zurich timezone

Contact

Session

AI/ML activities

4 Dec 2025, 09:00
31/3-004 - IT Amphitheatre (CERN)

31/3-004 - IT Amphitheatre

CERN

105
Show room on map

Conveners

AI/ML activities

  • Alessandro Di Girolamo (CERN)
  • Stefan Roiser (CERN)
  • Sofia Vallecorsa (CERN)
  • James Letts (Univ. of California San Diego (US))
  • Lorenzo Moneta (CERN)

Description

  • Multi-GPU benchmarks (examples such as ML Tracking per ATLAS, Reconstruction on * * HGCAL, FlashSim, Fastsim, MLPF, online)
  • Lessons learned: Setting up scalable infrastructure from NextGEN (Ricardo)
  • Looking ahead, e.g. Large Physics Models and Federated learning for HEP data

Presentation materials

There are no materials yet.

  1. 19. Introduction
    Lorenzo Moneta (CERN)
    04/12/2025, 09:00
  2. 22. ML - experience on inference and training in ATLAS
    Xiangyang Ju (Lawrence Berkeley National Lab. (US))
    04/12/2025, 09:10
  3. 23. Accelerating ML training and Hyperparameter Optimization with HPC
    Eric Wulff (CERN)
    04/12/2025, 09:50
  4. 7. Next Generation Triggers Work Package 1: feasible ML inference in grid production
    Axel Naumann (CERN)
    04/12/2025, 11:00
  5. 24. CERN ML Infrastructure
    Ricardo Rocha (CERN)
    04/12/2025, 11:20
  6. 25. Large Physics Models
    Sascha Caron (Nikhef National institute for subatomic physics (NL))
    04/12/2025, 11:50
  7. 20. Experience in Multi-GPU Training (ATLAS) TBC
  8. 21. Experience in Multi-GPU Training (CMS) TBC
Building timetable...