Communication efficient matrix multiplication on hypercubes

DOI10.1016/0167-8191(89)90091-4MaRDI QIDQ582005zbMATH OpenOpenAlexFDO

Authors Jarle Berntsen

Publication date 1989

Published in Parallel Computing (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/0167-8191(89)90091-4

zbMATH Keywords

asymptotic communication costblock algorithmscommunication overheadhypercubesmatrix multiplicationnearest neighbour communication

Mathematics Subject Classification ID

Parallel numerical computation (65Y05)

Review text

A method is presented for matrix multiplication on hypercubes, that uses only nearest neighbour communication and reduces the asymptotic communication cost.

Recommendations

• Communication-efficient matrix multiplication on hypercubes
• Communication lower bounds for distributed-memory matrix multiplication
• Communication and matrix computations on large message passing systems
• A method for exploiting communication/computation overlap in hypercubes
• Graph expansion and communication costs of fast matrix multiplication
• Matrix and vector operations on hypercube parallel processors
• Communication balancing in parallel sparse matrix-vector multiplication

Cited in

(4)

• Communication lower bounds and optimal algorithms for numerical linear algebra
• A RECURRENCE-FREE VARIANT OF STRASSEN'S ALGORITHM ON HYPERCUBE∗
• Pebbling Game and Alternative Basis for High Performance Matrix Multiplication
• Communication balancing in parallel sparse matrix-vector multiplication

This page was built for publication: Communication efficient matrix multiplication on hypercubes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q582005)