Poster 49: WarpX: Toward Exascale Modeling of Plasma Particle Accelerators on GPU

Authors: Maxence Thevenet (Lawrence Berkeley National Laboratory), Jean-Luc Vay (Lawrence Berkeley National Laboratory), Ann Almgren (Lawrence Berkeley National Laboratory), Diana Amorim (Lawrence Berkeley National Laboratory), John Bell (Lawrence Berkeley National Laboratory), Axel Huebl (Lawrence Berkeley National Laboratory), Revathi Jambunathan (Lawrence Berkeley National Laboratory), Rémi Lehe (Lawrence Berkeley National Laboratory), Andrew Myers (Lawrence Berkeley National Laboratory), Jaehong Park (Lawrence Berkeley National Laboratory), Olga Shapoval (Lawrence Berkeley National Laboratory), Weiqun Zhang (Lawrence Berkeley National Laboratory), Lixin Ge (SLAC National Accelerator Laboratory), Mark Hogan (SLAC National Accelerator Laboratory), Cho Ng (SLAC National Accelerator Laboratory), David Grote (Lawrence Livermore National Laboratory)

Abstract: Particle accelerators are a vital part of the DOE-supported infrastructure of discovery science and applications, but we need game-changing improvements in the size and cost for future accelerators. Plasma-based particle accelerators stand apart in their potential for these improvements. Turning this from a promising technology into mainstream scientific tools depends critically on high-performance, high-fidelity modeling of complex processes.

WarpX is an open-source particle-in-cell (PIC) code supported by the Exascale Computing Project (ECP) that is combining advanced algorithms with adaptive mesh refinement to allow challenging simulations of a multi-stage plasma-based TeV acceleration relevant for future high-energy physics discoveries. WarpX relies on the ECP co-design center for mesh refinement AMReX, and runs on CPU and GPU-accelerated computers. Production simulation have run on Cori KNL at NERSC and Summit at OLCF. In this poster, recent results and strategies on GPU will be presented, along with recent performance results.

Best Poster Finalist (BP): no

Poster: PDF
Poster summary: PDF


Back to Poster Archive Listing