Skip to Main content Skip to Navigation
Journal articles

Using Mixed Precision for Sparse Matrix Computations to Enhance the Performance while Achieving 64-bit Accuracy

Alfredo Buttari 1, 2 Jack Dongarra 3 Jakub Kurzak 3 Piotr Luszczek 3 Stanimir Tomov
1 IRIT-APO - Algorithmes Parallèles et Optimisation
IRIT - Institut de recherche en informatique de Toulouse
2 CNRS - Centre National de la Recherche Scientifique
3 ICL - Innovative Computing Laboratory [Knoxville]
Abstract : By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit accuracy of the resulting solution. These ideas can be applied to sparse multifrontal and supernodal direct techniques and sparse iterative techniques such as Krylov subspace methods. The approach presented here can apply not only to conventional processors but also to exotic technologies such as Field Programmable Gate Arrays (FPGA), Graphical Processing Units (GPU), and the Cell BE processor.
Document type :
Journal articles
Complete list of metadata
https://hal.archives-ouvertes.fr/hal-02421028
Contributor : Alfredo Buttari Connect in order to contact the contributor
Submitted on : Friday, December 20, 2019 - 11:23:47 AM
Last modification on : Tuesday, October 19, 2021 - 2:24:21 PM

Links full text

openaccess

Identifiers

Collections

UNIV-TLSE3 | UT1-CAPITOLE | UNIV-TLSE2 | CNRS | IRIT | IRIT-APO | IRIT-CISO | IRIT-CNRS

Citation

Alfredo Buttari, Jack Dongarra, Jakub Kurzak, Piotr Luszczek, Stanimir Tomov. Using Mixed Precision for Sparse Matrix Computations to Enhance the Performance while Achieving 64-bit Accuracy. ACM Transactions on Mathematical Software, Association for Computing Machinery, 2008, 34 (4), pp.1-22. ⟨10.1145/1377596.1377597⟩. ⟨hal-02421028⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Les métriques sont temporairement indisponibles