English

Practical Implementation of Lattice QCD Simulation on SIMD Machines with Intel AVX-512

High Energy Physics - Lattice 2018-11-05 v1 Computational Physics

Abstract

We investigate implementation of lattice Quantum Chromodynamics (QCD) code on the Intel AVX-512 architecture. The most time consuming part of the numerical simulations of lattice QCD is a solver of linear equation for a large sparse matrix that represents the strong interaction among quarks. To establish widely applicable prescriptions, we examine rather general methods for the SIMD architecture of AVX-512, such as using intrinsics and manual prefetching, for the matrix multiplication. Based on experience on the Oakforest-PACS system, a large scale cluster composed of Intel Xeon Phi Knights Landing, we discuss the performance tuning exploiting AVX-512 and code design on the SIMD architecture and massively parallel machines. We observe that the same code runs efficiently on an Intel Xeon Skylake-SP machine.

Cite

@article{arxiv.1811.00893,
  title  = {Practical Implementation of Lattice QCD Simulation on SIMD Machines with Intel AVX-512},
  author = {Issaku Kanamori and Hideo Matsufuru},
  journal= {arXiv preprint arXiv:1811.00893},
  year   = {2018}
}

Comments

17 pages, 9 figures, talk given by I.K. at the Workshop Large Scale Computational Physics (LSCP 2018) in the 18th International Conference on Computational Science and its Applications (ICCSA 2018), 2-5 July 2018, Melbourne. arXiv admin note: text overlap with arXiv:1712.01505

R2 v1 2026-06-23T05:02:10.768Z