Presentation
DBSR: An Efficient Storage Format for Vectorizing Sparse Triangular Solvers on Structured Grids
SessionHigh-Performance Solvers
DescriptionThe Sparse Triangular Solver (SPTRSV) plays a critical role in solving structured grid problems. Yet, the commonly used sparse matrix storage formats for structured grid methods do not efficiently support SPTRSV in utilizing the instruction parallelism offered by modern multi-core CPUs. We introduce DBSR, a new sparse storage format to enable SPTRSV to take advantage of the SIMD instructions. DBSR promotes contiguous memory access and vectorized computation, while also optimizing memory usage. We evaluate DBSR by applying it within multigrid algorithms and the zero fill-in incomplete LU preconditioner. Our evaluation, conducted on four architectures -- three ARMv8 systems and one x86 system -- demonstrates that DBSR consistently outperforms mainstreamed storage formats across evaluation workloads and platforms.
Event Type
Paper
TimeWednesday, 20 November 20244pm - 4:30pm EST
LocationB312-B313A
Accelerators
Algorithms
Linear Algebra
Modeling and Simulation
Numerical Methods
TP
Archive
view