Exascale supercomputers such as Frontier enable WarpX simulations of novel concepts for the compact generation and acceleration of high-charge electron beams with a novel hybrid solid-gas target concept. At left, a laser pulse (red and blue) creates a large amplitude electric wakefield (orange and yellow) as it propagates through a low-density plasma. In the center image, after reflection on a high-density target, many electrons (red dots) are extracted from the laser and accelerated by the electric wakefield. At right, the extracted electrons that are trapped by the wakefield form a high-charge electron beam that follows the laser and is accelerated to very high energy in a very short distance.

Exascale computing enables the exploration of particle accelerator designs over a much larger range of configurations, leading to more energetic and denser beams of higher quality than would be possible otherwise. It also eases research on novel accelerator technologies, such as plasma-based accelerators, that are especially demanding computationally. It allows for modeling fusion plasmas and devices on a much wider range of space and time scales, leading to more accurate and predictive simulations and improved designs.

In exascale computing, as in our other endeavors, we are active partners in, and early adopters of, performance-portable algorithm design and implementation, petabyte-scale data I/O, research into AI/ML methods, autonomous optimization, scalable in situ algorithms and visualization, workflow development, usability, benchmarking and performance tuning, and sustainable open software engineering.

Berkeley Lab is home to invaluable central resources for modeling and simulation in a strong and highly collaborative Computing Sciences Area that includes the National Energy Research Supercomputing Center.

An example of the results of this synergy is WarpX, a community open-source Particle-In-Cell simulation code developed by an international collaboration under AMP’s leadership with the support of the DOE Exascale Computing Project.  The achievement was recognized with the 2022 ACM Gordon Bell Prize and the 2023 Berkeley Lab Director’s Award for Scientific Achievement.