Heterogeneous reconstruction of hadronic particle flow clusters with the Alpaka Portability Library
In response to increasing data challenges, CMS has adopted the use of GPU offloading at the High-Level Trigger (HLT). However, GPU code is often hardware specific, and increases the maintenance burden on software development. The Alpaka (Abstraction Library for Parallel Kernel Acceleration) portability library offers a solution to this issue, and has been implemented into the CMS software (CMSSW) for use online at HLT. A portion of the final-state particle candidate reconstruction algorithm, Particle Flow, represented a target for increased performance through parallel operation. We discuss the port of hadronic Particle Flow clustering to Alpaka, and the validation of physics and performance at HLT for 2024 data taking.
Authors
- F. Pantaleo (ORCID: https://orcid.org/0000-0003-3266-4357)
- K. Hatakeyama (ORCID: https://orcid.org/0000-0002-6012-2451)
- W. Redjeb (ORCID: https://orcid.org/0000-0001-9794-8292)
- A. Bocci (ORCID: https://orcid.org/0000-0002-6515-5666)
- Florian Lorkowski (ORCID: https://orcid.org/0000-0003-2677-3805)
- A. Das (ORCID: https://orcid.org/0000-0001-9115-9698)
- M. Saunders (ORCID: https://orcid.org/0000-0003-1572-9075)
- Jonathan Samudio (ORCID: https://orcid.org/0000-0002-4767-8463)
Institutions
- University of Notre Dame (US)
- Baylor University (US)
- Deutsches Elektronen-Synchrotron DESY (DE)
- European Organization for Nuclear Research (CH)
- RWTH Aachen University (DE)
Publication Details
- Journal
- EPJ Web of Conferences
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1051/epjconf/202533701171
- Primary Topic
- Particle physics theoretical and experimental studies
- Type
- article
- Field-Weighted Citation Impact
- 0.00