The work proposes ffMDF, a lightweight dynamic run-time support able to achieve high performance in the execution of dense linear algebra kernels on shared-cache multi-core. ffMDF implements a dynamic macro-dataflow interpreter processing DAG graphs generated on-the-fly out of standard numeric kernel code. The experimental results demonstrate that the performance obtained using ffMDF on both fine-grain and coarse-grain problems is comparable with or even better than that achieved by de-facto standard so- lutions (notably PLASMA library), which use separate run-time supports specifically optimised for different computational grains on modern multi-core.
A Lightweight Run-Time Support for Fast Dense Linear Algebra on Multi-CoreSoftware Engineering
BUONO, DANIELE;DANELUTTO, MARCO;DE MATTEIS, TIZIANO;MENCAGLI, GABRIELE;TORQUATI, MASSIMO
2014-01-01
Abstract
The work proposes ffMDF, a lightweight dynamic run-time support able to achieve high performance in the execution of dense linear algebra kernels on shared-cache multi-core. ffMDF implements a dynamic macro-dataflow interpreter processing DAG graphs generated on-the-fly out of standard numeric kernel code. The experimental results demonstrate that the performance obtained using ffMDF on both fine-grain and coarse-grain problems is comparable with or even better than that achieved by de-facto standard so- lutions (notably PLASMA library), which use separate run-time supports specifically optimised for different computational grains on modern multi-core.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
