Many-gluon tree amplitudes on modern GPUs: A case study for novel event generators
Enrico Bothmann, Walter Giele, Stefan Hoeche, Joshua Isaacson, Max Knobbe
SciPost Phys. Codebases 3 (2022) · published 23 August 2022
- doi: 10.21468/SciPostPhysCodeb.3
- live repo (external)
- Submissions/Reports
-
This Publication is part of a bundle
When citing, cite all relevant items (e.g. for a Codebase, cite both the article and the release you used).
| DOI | Type | |
|---|---|---|
| 10.21468/SciPostPhysCodeb.3 | Article | |
| 10.21468/SciPostPhysCodeb.3-r1.0 | Codebase release |
Abstract
The compute efficiency of Monte-Carlo event generators for the Large Hadron Collider is expected to become a major bottleneck for simulations in the high-luminosity phase. Aiming at the development of a full-fledged generator for modern GPUs, we study the performance of various recursive strategies to compute multi-gluon tree-level amplitudes. We investigate the scaling of the algorithms on both CPU and GPU hardware. Finally, we provide practical recommendations as well as baseline implementations for the development of future simulation programs. The GPU implementations can be found at: https://www.gitlab.com/ebothmann/blockgen-archive.
Cited by 8
Authors / Affiliations: mappings to Contributors and Organizations
See all Organizations.- 1 Enrico Bothmann,
- 2 Walter Giele,
- 2 Stefan Höche,
- 2 Joshua Isaacson,
- 1 Max Knobbe
- 1 Georg-August-Universität Göttingen / University of Göttingen [GAU]
- 2 Fermi National Accelerator Laboratory [Fermilab]