关键词: 可伸缩矢量扩展, 光滑粒子流体动力学, 向量化

Abstract: Smoothed Particle Hydrodynamics (SPH) is a meshless particle method that has emerged in recent years. SPH has obvious advantages in dealing with large deformations, moving material surfaces, and free surfaces. In recent years, it has been widely used in the field of numerical simulation, and it is a typical application of scientific computing. As an explicit particle method, SPH needs to calculate a large number of particle interactions in each iteration step, and the amount of calculation is very large. How to improve the calculation efficiency of SPH has become a research hotspot. Scalable Vector Extension (SVE) is the next generation SIMD extension proposed by ARM. This paper proposes a SIMD optimization method for SPH based on SVE, and achieves a good speedup.

Key words: scalable vector extension(SVE), smoothed particle hydrodynamics, vectorization