Co-design of a particle-in-cell plasma simulation code for Intel Xeon Phi: a first look at Knights Landing

TL;DR

This paper evaluates and optimizes a particle-in-cell plasma simulation code for Intel Xeon Phi Knights Landing, achieving significant speedups and demonstrating the benefits of code co-design across different architectures.

Contribution

It presents the first performance results of PICADOR on Knights Landing and demonstrates effective code optimizations that improve performance across multiple architectures.

Findings

2.43x speedup from Knights Corner to Knights Landing

Additional 1.89x speedup from code optimization

Achieved 100 GFLOPS double precision performance on Knights Landing

Abstract

Three dimensional particle-in-cell laser-plasma simulation is an important area of computational physics. Solving state-of-the-art problems requires large-scale simulation on a supercomputer using specialized codes. A growing demand in computational resources inspires research in improving efficiency and co-design for supercomputers based on many-core architectures. This paper presents first performance results of the particle-in-cell plasma simulation code PICADOR on the recently introduced Knights Landing generation of Intel Xeon Phi. A straightforward rebuilding of the code yields a 2.43 x speedup compared to the previous Knights Corner generation. Further code optimization results in an additional 1.89 x speedup. The optimization performed is beneficial not only for Knights Landing, but also for high-end CPUs and Knights Corner. The optimized version achieves 100 GFLOPS double precision performance on a Knights Landing device with the speedups of 2.35 x compared to a 14-core Haswell CPU and 3.47 x compared to a 61-core Knights Corner Xeon Phi.