Title :
Solving tridiagonal systems on a GPU
Author :
Murphy, Brian J.
Author_Institution :
Dept. of Math. & Comput. Sci., City Univ. of New York, Bronx, NY, USA
Abstract :
We implement a parallel tridiagonal solver based on cyclic reduction (CR) for a graphics processing unit (GPU). The bane of such solvers is a low computation to communication ratio. With this our main consideration we focus our effort on lowering communication costs. In so doing we accelerate system solving. Further, in the diagonally dominant case computation is decoupled into independent partitions allowing for efficient processing of larger systems.
Keywords :
cost reduction; graphics processing units; parallel processing; CR; GPU; communication costs; cyclic reduction; diagonally dominant case computation; graphics processing unit; larger system processing; parallel tridiagonal solver; tridiagonal systems; Equations; Graphics processing units; Instruction sets; Registers; Sparse matrices; Vectors; Cyclic Reduction; Diagonal Dominance; Diagonally Dominant; GPU; Graphics Processing Unit; LU Decomposition; Linear System Solver; Tridiagonal;
Conference_Titel :
High Performance Computing (HiPC), 2013 20th International Conference on
Conference_Location :
Bangalore
DOI :
10.1109/HiPC.2013.6799117
