Using Static Allocation Algorithms for Matrix Matrix Multiplication on Multicores and GPUs - Archive ouverte HAL Access content directly
Conference Papers Year :

Using Static Allocation Algorithms for Matrix Matrix Multiplication on Multicores and GPUs

(1, 2) , (1, 3)
1
2
3

Abstract

We consider the problem of data allocation when performing matrix multiplication on a heterogeneous node, with multicores and GPUs. Classical (cyclic) allocations designed for homogeneous settings are not appropriate, but the advent of task-based runtime systems makes it possible to use more general allocations. Previous theoretical work has proposed square and cube partitioning algorithms aimed at minimizing data movement for matrix multiplication. We propose techniques to adapt these continuous square partitionings to allocating discrete tiles of a matrix, and strategies to adapt the static allocation at run-time. We use these techniques in an implementation of Matrix Multiplication based on the StarPU runtime system, and we show through extensive experiments that this implementation allows to consistently obtain a lower communication volume while improving slightly the execution time, compared to standard state-of-the-art dynamic strategies.
Fichier principal
Vignette du fichier
icpp.pdf (747.23 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01670678 , version 1 (22-12-2017)
hal-01670678 , version 2 (31-05-2018)

Identifiers

Cite

Lionel Eyraud-Dubois, Thomas Lambert. Using Static Allocation Algorithms for Matrix Matrix Multiplication on Multicores and GPUs. ICPP 2018 - 47th International Conference on Parallel Processing, Aug 2018, Eugene, OR, United States. ⟨10.1145/3225058.3225066⟩. ⟨hal-01670678v2⟩
256 View
391 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More