Concurrency Practice and Experience · 1994 · 153 citations · 10 references
ComputingComputer EngineeringComputer ScienceAlgorithmsParallel AlgorithmsParallel ComputingDistributed ComputingDistributed Memory ParallelismDistributed Memory SystemsMatrix ComputationParallel ArchitecturesParallel Computer ArchitectureDistributed Linear AlgebraDistributed MemoryDistributed Memory AlgorithmsMatrix MultiplicationParallel Matrix AlgorithmsParallel Matrix ComputationParallel Matrix MultiplicationParallel Matrix-vector Multiplication
Abstract The paper describes Parallel Universal Matrix Multiplication Algorithms (PUMMA) on distributed memory concurrent computers. The PUMMA package includes not only the non‐transposed matrix multiplication routine C = A ⋅ B, but also transposed multiplication routines C = A T ⋅ B, C = A ⋅ B T , and C = A T ⋅ B T , for a block cyclic data distribution. The routines perform efficiently for a wide range of processor configurations and block sizes. The PUMMA together provide the same functionality as the Level 3 BLAS routine xGEMM. Details of the parallel implementation of the routines are given, and results are presented for runs on the Intel Touchstone Delta computer.
10
A set of level 3 basic linear algebra subprograms
Jack Dongarra, Jeremy Du Croz, Sven Hammarling et al. · ACM Transactions on Mathematical Software · 1990 · 1.8K citations · Full text
Solving Linear Systems on Vector and Shared Memory Computers.
James M. Ortega, Jack Dongarra, Iain Duff et al. · Mathematics of Computation · 1992 · 439 citations
A look at scalable dense linear algebra libraries
Jack Dongarra, Robert A. Geijn, David Walker · 2003 · 69 citations