DocumentCode :
2347252
Title :
Empirically tuning LAPACK’s blocking factor for increased performance
Author :
Whaley, R. Clint
Author_Institution :
Dept. of Comput. Sci., Univ. of Texas at San Antonio, San Antonio, TX
fYear :
2008
fDate :
20-22 Oct. 2008
Firstpage :
303
Lastpage :
310
Abstract :
LAPACK (linear algebra package) is a statically cache-blocked library, where the blocking factor (NB) is determined by the service routine ILAENV. Users are encouraged to tune NB to maximize performance on their platform/BLAS (the BLAS are LAPACKpsilas computational engine), but in practice very few users do so (both because it is hard, and because its importance is not widely understood). In this paper we (1) Discuss our empirical tuning framework for discovering good NB settings, (2) quantify the performance boost that tuning NB can achieve on several LAPACK routines across multiple architectures and BLAS implementations, (3) compare the best performance of LAPACKpsilas statically blocked routines against state of the art recursively blocked routines, and vendor-optimized LAPACK implementations, to see how much performance loss is mandated by LAPACKpsilas present static blocking strategy, and finally (4) use results to determine how best to block nonsquare matrices once good square blocking factors are discovered.
Keywords :
cache storage; mathematics computing; matrix decomposition; optimisation; recursive functions; software libraries; software packages; ILAENV service routine; LAPACK linear algebra package; basic linear algebra subprogram; block nonsquare matrix; matrix factorization; recursively blocked routine; square blocking factor; static blocking strategy; statical cache-blocked library; vendor-optimized LAPACK; Computational efficiency; Computer architecture; Computer science; Costs; Information technology; Libraries; Linear algebra; Niobium; Packaging machines; Performance loss;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Science and Information Technology, 2008. IMCSIT 2008. International Multiconference on
Conference_Location :
Wisia
Print_ISBN :
978-83-60810-14-9
Type :
conf
DOI :
10.1109/IMCSIT.2008.4747256
Filename :
4747256
Link To Document :
بازگشت