Parallel and heterogeneous m-Hessenberg-triangular-triangular reduction (CROSBI ID 235447)
Prilog u časopisu | izvorni znanstveni rad | međunarodna recenzija
Podaci o odgovornosti
Bosner, Nela ; Karlsson, Lars
engleski
Parallel and heterogeneous m-Hessenberg-triangular-triangular reduction
The m-Hessenberg-triangular-triangular (mHTT) reduction is a simultaneous orthogonal reduction of three matrices to condensed form. It has applications, for example, in solving shifted linear systems arising in various control theory problems. A new heterogeneous CPU/GPU implementation of the mHTT reduction is presented and evaluated against an existing CPU implementation. The algorithm offloads the compute-intensive matrix-matrix multiplications to the GPU and keeps the inner loop, which is memory intensive and has a complicated control flow, on the CPU. Experiments demonstrate that the heterogeneous implementation can be superior to the existing CPU implementation on a system with 2x8 CPU cores and one GPU. Future development should focus on improving the scalability of the CPU computations.
m-Hessenberg-triangular-triangular form, solving shifted linear systems, Givens rotations, heterogeneous CPU/GPU implementation
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o izdanju
39 (1)
2017.
C29-C47
objavljeno
1064-8275
1095-7197
10.1137/15M1047349