RT info:eu-repo/semantics/article
T1 Auto-tuned nested parallelism: A way to reduce the execution time of scientific software in NUMA systems
A1 Cámara Moreno, Jesús
A1 Cuenca, Javier
A1 García, Luis Pedro
A1 Giménez, Domingo
K1 Computación Paralela
K1 Auto-Tuning
K1 Auto-tuning
K1 Linear Algebra
K1 Performance Modeling
K1 1203 Ciencia de Los Ordenadores
K1 3304 Tecnología de Los Ordenadores
AB The most computationally demanding scientific problems are solved with large parallel systems. In some cases these systems are Non-Uniform Memory Access (NUMA) multiprocessors made up of a large number of cores which share a hierarchically organized memory. The main basic component of these scientific codes is often matrix multiplication, and the efficient development of other linear algebra packages is directly based on the matrix multiplication routine implemented in the BLAS library. BLAS library is used in the form of packages implemented by the vendors or free implementations. The latest versions of this library are multithreaded and can be used efficiently in multicore systems, but when they are used inside parallel codes, the two parallelism levels can interfere and produce a degradation of the performance. In this work, an auto-tuning method is proposed to select automatically the optimum number of threads to use at each parallel level when multithreaded linear algebra routines are called from OpenMP parallel codes. The method is based on a simple but effective theoretical model of the execution time of the two-level routines. The methodology is applied to a two-level matrix–matrix multiplication and to different matrix factorizations (LU, QR and Cholesky) by blocks. Traditional schemes which directly use the multithreaded routine of BLAS, dgemm, are compared with schemes combining the multithreaded dgemm with OpenMP.
PB Elsevier
SN 0167-8191
YR 2014
FD 2014
LK https://uvadoc.uva.es/handle/10324/74462
UL https://uvadoc.uva.es/handle/10324/74462
LA eng
NO Parallel Computing, 2014, Volume 40, Issue 7, Pages 309-327
NO Producción Científica
DS UVaDOC
RD 04-abr-2025