Advanced Vectorization of PPML Method for Intel® Xeon® Scalable Processors
Piecewise Parabolic Method on a Local Stencil is very useful for numerical simulation of fluid dynamics, astrophysics. The main idea of the PPML method is the use of a piecewise parabolic numerical solution on the previous time step for computing the Riemann problem solving partial differential equations system (PDE). In this paper, we present the new version of PDE solver which is based on the PPML method optimized for Intel Xeon Scalable processor family. The results of performance comparison between different types of AVX-512 compatible Intel Xeon Scalable processors are presented. Special attention is paid to comparing the performance of Intel Xeon Phi (KNL) and Intel Xeon Scalable processors.
KeywordsMassively parallel supercomputers Astrophysics Code vectorization
This work was partially supported by RFBR grants 18-07-00757, 18-01-00166 and 16-07-00434. Methodical work was partially supported by the Grant of the Russian Science Foundation grant 16-11-10028.
- 1.Vectorization: A Key Tool To Improve Performance On Modern CPUs. https://software.intel.com/en-us/articles/vectorization-a-key-tool-to-improve-performance-on-modern-cpus
- 2.Intel Xeon Scalable Debuts. https://hothardware.com/reviews/intel-xeon-scalable-processor-family-review?page=2
- 3.Vshivkov, V.A., Lazareva, G.G., Snytnikov, A.V., Kulikov, I.M., Tutukov, A.V.: Hydrodynamical code for numerical simulation of the gas components of colliding galaxies. Astrophys. J. Suppl. Ser. 194(47), 1–12 (2011)Google Scholar
- 7.Chernykh, I., Stoyanovskaya, O., Zasypkina, O.: ChemPAK software package as an environment for kinetics scheme evaluation. Chem. Prod. Process Model. 4(4) (2009)Google Scholar
- 11.Intel Advisor. https://software.intel.com/en-us/intel-advisor-xe
- 13.Glinskiy, B., Kulikov, I., Chernykh, I.: Improving the performance of an AstroPhi code for massively parallel supercomputers using roofline analysis. Commun. Comput. Inf. Sci. 793, 400–406 (2017)Google Scholar
- 14.Siberian Supercomputer Center ICMMG SB RAS. http://www2.sscc.ru