TY - ECHAP
AU - M. Karwacki
AU - Przemysław Stpiczyński
AU - K. De Bosschere
AU - E.H. D`Hollander
AU - G.R. Joubert
AU - D. Padna
AU - F. Peters
AU - M. Sawyer
AB - CUBLAS is a widely used implementation of BLAS (Basic Linear Algebra Subprograms) for NVIDIA CUDA Graphical Processing Units (GPUs). The aim of this paper is to show that the performance of the selected Level 2 BLAS routines for working with triangular matrices can be improved using some optimization techniques suitable for GPUs like using shared memory and coalesced memory access. We present new implementation of the routines xTRMV and xTRSV. The results of experiments carried out on two GPU architectures: Tesla M2050 and GeForce GTX 260 show that these new implementations are up to 500% faster than corresponding routines from CUBLAS Library.
BT - Advances in Parallel Computing
LA - eng
N1 - DOI: 10.3233/978-1-61499-041-3-405
N2 - CUBLAS is a widely used implementation of BLAS (Basic Linear Algebra Subprograms) for NVIDIA CUDA Graphical Processing Units (GPUs). The aim of this paper is to show that the performance of the selected Level 2 BLAS routines for working with triangular matrices can be improved using some optimization techniques suitable for GPUs like using shared memory and coalesced memory access. We present new implementation of the routines xTRMV and xTRSV. The results of experiments carried out on two GPU architectures: Tesla M2050 and GeForce GTX 260 show that these new implementations are up to 500% faster than corresponding routines from CUBLAS Library.
PB - IOS Press
PY - 2012
SE - Improving performance of triangular Matrix-Vector BLAS routines on GPUs.
SN - 978-1-61499-040-6
T2 - Advances in Parallel Computing
TI - Applications and Techniques on the Road to Exascale Computing
VL - 22
ER -