Skip to Main content Skip to Navigation
Reports

On analysis of partitioning models and metrics in parallel sparse matrix-vector multiplication

Abstract : Graph/hypergraph partitioning models and methods have been successfully used to minimize the communication requirements among processors in several parallel computing applications. Parallel sparse matrix-vector multiplication~(SpMxV) is one of the representative applications that renders these models and methods indispensable in many scientific computing contexts. We investigate the interplay of several partitioning metrics and execution times of SpMxV implementations in three libraries: Trilinos, PETSc, and an in-house one. We design and carry out experiments with up to 512 processors and investigate the results with regression analysis. Our experiments show that the partitioning metrics, although not an exact measure of communication cost, influence the performance greatly in a distributed memory setting. The regression analyses demonstrate which metric is the most influential for the execution time of the three libraries used.
Complete list of metadatas

https://hal.inria.fr/hal-00821523
Contributor : Equipe Roma <>
Submitted on : Friday, May 10, 2013 - 4:10:22 PM
Last modification on : Sunday, September 29, 2019 - 9:58:17 PM
Long-term archiving on: : Tuesday, April 4, 2017 - 6:13:16 AM

File

RR-8301.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00821523, version 1

Citation

Umit Catalyurek, Kamer Kaya, Bora Uçar. On analysis of partitioning models and metrics in parallel sparse matrix-vector multiplication. [Research Report] RR-8301, 2013, pp.25. ⟨hal-00821523v1⟩

Share

Metrics

Record views

239

Files downloads

44