Performance and energy consumption of accurate and mixed-precision linear algebra kernels on GPUs

DOI10.1016/J.CAM.2019.112701zbMATH Open1493.65282OpenAlexW2999652941WikidataQ126398654 ScholiaQ126398654MaRDI QIDQ2297181FDOQ2297181

Authors: Daichi Mukunoki, Takeshi Ogita 

Publication date: 18 February 2020

Published in: Journal of Computational and Applied Mathematics (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/j.cam.2019.112701

Recommendations

• Mixed precision algorithms in numerical linear algebra
• Mixed precision block fused multiply-add: error analysis and application to GPU tensor cores
• Mixed-precision iterative refinement using tensor cores on GPUs to accelerate solution of linear systems
• Accelerating GPU kernels for dense linear algebra
• Design, implementation and testing of extended and mixed precision BLAS

zbMATH Keywords

BLASenergy consumptionaccuracyGPUSpMVmixed-precision

Mathematics Subject Classification ID

Numerical algorithms for specific classes of architectures (65Y10) Numerical linear algebra (65Fxx)

Cites Work

• The University of Florida sparse matrix collection
• MPFR
• Basic Linear Algebra Subprograms for Fortran Usage
• Title not available (Why is that?)
• Accurate Sum and Dot Product
• A floating-point technique for extending the available precision
• Design, implementation and testing of extended and mixed precision BLAS
• High-precision division and square root
• Accelerating the solution of linear systems by iterative refinement in three precisions
• Reproducible and accurate matrix multiplication

Cited In (5)

• Infinite-precision inner product and sparse matrix-vector multiplication using Ozaki scheme with Dot2 on manycore processors
• Matrix Multiplication in Multiword Arithmetic: Error Analysis and Application to GPU Tensor Cores
• Mixed precision block fused multiply-add: error analysis and application to GPU tensor cores
• GPU Based Mixed Precision PWR Depletion Calculation
• Mixed-precision conjugate gradient algorithm using the groupwise update strategy

Uses Software

• BLAS
• MPFR
• SparseMatrix
• LBNL
• XBLAS
• MPACK
• ExBLAS
• ReproBLAS

This page was built for publication: Performance and energy consumption of accurate and mixed-precision linear algebra kernels on GPUs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2297181)