Parallel experiments with RARE-BLAS - Archive ouverte HAL Access content directly
Conference Papers Year : 2016

Parallel experiments with RARE-BLAS

(1) , (1) , (1)
1

Abstract

Numerical reproducibility failures rise in parallel computation because of the non-associativity of floating-point summation. Optimizations on massively parallel systems dynamically modify the floating-point operation order. Hence, numerical results may change from one run to another. We propose to ensure reproducibility by extending as far as possible the IEEE-754 correct rounding property to larger operation sequences. Our RARE-BLAS (Reproducible, Accurately Rounded and Efficient BLAS) benefits from recent accurate and efficient summation algorithms. Solutions for level 1 (asum, dot and nrm2) and level 2 (gemv) routines are provided. We compare their performance to the Intel MKL library and to other existing reproducible algorithms. For both shared and distributed memory parallel systems, we exhibit an extra-cost of 2× in the worst case scenario, which is satisfying for a wide range of applications. For Intel Xeon Phi accelerator a larger extra-cost (4× to 6×) is observed, which is still helpful at least for debugging and validation.
Fichier principal
Vignette du fichier
SYNASC.pdf (197.39 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

lirmm-01349698 , version 1 (28-07-2016)

Identifiers

Cite

Chemseddine Chohra, Philippe Langlois, David Parello. Parallel experiments with RARE-BLAS. SYNASC: Symbolic and Numeric Algorithms for Scientific Computing, Sep 2016, Timisoara, Romania. pp.135-138, ⟨10.1109/SYNASC.2016.032⟩. ⟨lirmm-01349698⟩
224 View
410 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More