10–15 Aug 2025
SAFE Credit Union Convention Center
America/Los_Angeles timezone

Towards Differentiable Beam Dynamics Modeling in BLAST/ImpactX

TUP101
12 Aug 2025, 16:00
2h
SAFE Credit Union Convention Center

SAFE Credit Union Convention Center

1401 K St, Sacramento, CA 95814
Poster Presentation MC5 – Beam Dynamics and EM Fields TUP: Tuesday Poster Session

Speaker

Axel Huebl (Lawrence Berkeley National Laboratory)

Description

Differentiable simulations are in demand in accelerator physics, demonstrating order-of-magnitude improvements for complex tasks such as many-parameter optimization for accelerator working points and reconstruction of hard-to-measure quantities. At its core, a differentiable simulation does not only solve a forward problem, but additionally provides gradients of output parameters (e.g. beam parameters) with respect to input parameters (e.g. beamline or source parameters).
How to effectively program large dynamic simulations differentiably is still an open question, but there is general consensus that a “single-source” approach aided by automatic differentiation (AD) is desirable. Addressing this, there are a) emerging domain-specific languages in machine learning that are intrinsically differentiable, and b) highly-performing & scalable, general-purpose languages like ISO C++ of existing codes. The challenge of approach a) is syntax specialization, which can limit ease of implementation & performance for physics algorithms, while b) requires additional work for AD.
Performance is important for modeling high-order beam dynamics and collective effects in accelerators. We compare the fast, modern codes ImpactX (C++/Python) and Cheetah (PyTorch) using traditional, gradient-free modeling. We then show progress in introducing single-source differentiability in ImpactX using modern compiler techniques, producing performant executables for gradient-based and gradient-free modeling.

Funding Agency

Supported by the CAMPA collaboration, a project of the US DOE Office of Science, ASCR & HEP, SciDAC program. Seeded by the LBNL LDRD program under US DOE contract DE-AC02-05CH11231.

Footnotes

  • Huebl A et al., NAPAC22, TUYE2 (2022) DOI:10.18429/JACoW-NAPAC2022-TUYE2
    ** Kaiser J et al., PRAB 27, 054601 (2024) DOI:10.1103/PhysRevAccelBeams.27.054601
    *** Moses W et al., NEURIPS2020, SC21, SC22 (2020-2022) DOI:10.1145/3458817.3476165 DOI:10.5555/3571885.3571964
I have read and accept the Privacy Policy Statement Yes
Please consider my poster for contributed oral presentation No
Would you like to submit this poster in student poster session on Sunday (August 10th) No

Author

Axel Huebl (Lawrence Berkeley National Laboratory)

Co-authors

Chad Mitchell (Lawrence Berkeley National Laboratory) Remi Lehe (Lawrence Berkeley National Laboratory) Mr Grégoire Charleux (Lawrence Berkeley National Laboratory) Dr Andrew Myers (Lawrence Berkeley National Laboratory) Dr Weiqun Zhang (Lawrence Berkeley National Laboratory) Ji Qiang (Lawrence Berkeley National Laboratory) Jean-Luc Vay (Lawrence Berkeley National Laboratory) Jan Kaiser (Deutsches Elektronen-Synchrotron DESY) Mr Christian Hespe (Deutsches Elektronen-Synchrotron DESY) Juan Pablo Gonzalez-Aguilera (University of Chicago) Chenran Xu (Argonne National Laboratory) Dr Andrea Santamaria Garcia (University of Liverpool) Ryan Roussel (SLAC National Accelerator Laboratory) Auralee Edelen (SLAC National Accelerator Laboratory) Prof. William Steven Moses (University of Illinois Urbana-Champaign)

Presentation materials

There are no materials yet.