Inline vector compression for computational physics

DOI10.1016/J.CPC.2020.107562arXiv2003.02633OpenAlexW3081894855MaRDI QIDQ6041006FDOQ6041006

Authors: W. Trojak, F. D. Witherden 

Publication date: 25 May 2023

Published in: Computer Physics Communications (Search for Journal in Brave)

Abstract: A novel inline data compression method is presented for single-precision vectors in three dimensions. The primary application of the method is for accelerating computational physics calculations where the throughput is bound by memory bandwidth. The scheme employs spherical polar coordinates, angle quantisation, and a bespoke floating-point representation of the magnitude to achieve a fixed compression ratio of 1.5. The anisotropy of this method is considered, along with companding and fractional splitting techniques to improve the efficiency of the representation. We evaluate the scheme numerically within the context of high-order computational fluid dynamics. For both the isentropic convecting vortex and the Taylor--Green vortex test cases, the results are found to be comparable to those without compression. Performance is evaluated for a vector addition kernel on an NVIDIA Titan V GPU; it is demonstrated that a speedup of 1.5 can be achieved.

Full work available at URL: https://arxiv.org/abs/2003.02633

zbMATH Keywords

flux reconstructionGPU computingvector compression

Mathematics Subject Classification ID

Numerical analysis (65-XX) Fluid mechanics (76-XX)

Cites Work

• PyFR: an open source framework for solving advection-diffusion type problems on streaming architectures using the flux reconstruction approach
• Compression of individual sequences via variable-rate coding
• A universal algorithm for sequential data compression
• A high-order accurate discontinuous finite element method for the numerical solution of the compressible Navier-Stokes equations
• Small-scale structure of the Taylor–Green vortex
• A new class of high-order energy stable flux reconstruction schemes
• Low-storage, explicit Runge-Kutta schemes for the compressible Navier-Stokes equations
• Simplified Second-Order Godunov-Type Methods
• Impact of the floating-point precision and interpolation scheme on the results of DNS of turbulence by pseudo-spectral codes
• ZEFR: a GPU-accelerated high-order solver for compressible viscous flows using the flux reconstruction method

Cited In (1)

• Hyperbolic diffusion in flux reconstruction: optimisation through kernel fusion within tensor-product elements

This page was built for publication: Inline vector compression for computational physics

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6041006)