A new whitepaper looks at Accelerated Sparse Triangular Linear Systems:
“Fresh from NVIDIA Numeric Libraries Team, a white paper illustrating the use the CUSPARSE and CUBLAS libraries to achieve a 2x speed up of incomplete-LU and Cholesky preconditioned interative methods. Focusing on the Bi-Conjugate Gradient, Stabilized and Conjugate Gradient iterative methods that can be used to solve large sparse nonsymmetric and symmetric positive de-nite linear systems, respectively. Also, we comment on the parallel sparse triangular solve, which is an essential building block in these algorithms.
Read the Full Story.