Analytical Modeling Is Enough for High-Performance BLIS
ACM Transactions on Mathematical Software2016Vol. 43(2), pp. 1–18
Citations Over TimeTop 1% of 2016 papers
Abstract
We show how the BLAS-like Library Instantiation Software (BLIS) framework, which provides a more detailed layering of the GotoBLAS (now maintained as OpenBLAS) implementation, allows one to analytically determine tuning parameters for high-end instantiations of the matrix-matrix multiplication. This is of both practical and scientific importance, as it greatly reduces the development effort required for the implementation of the level-3 BLAS while also advancing our understanding of how hierarchically layered memories interact with high-performance software. This allows the community to move on from valuable engineering solutions (empirically autotuning) to scientific understanding (analytical insight).
Related Papers
- → On optimizing multiplications of sparse matrices(1996)9 cited
- → Approximate Weighted C R Coded Matrix Multiplication(2021)2 cited
- → Large Scale 3D Multi-Phase-Field Simulation of Microstructure Evolution Using TSUBAME2.5 GPU-Supercomputer(2014)1 cited
- → On the Multiplication of Biquaternions(2015)1 cited