Machine learning approach for the predicting performance of SpMV on GPU

Akrem Benatia, Weixing Ji, Yizhuo Wang, Feng Shi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

15 Citations (Scopus)

Abstract

Sparse Matrix-Vector Multiplication (SpMV) kernel dominates the computing cost in numerous scientific applications. Many implementations based on different sparse formats were proposed recently for optimizing this kernel on the GPU side. Since the performance of the SpMV varies significantly according to the sparsity characteristics of the input matrix and the hardware features, developing an accurate performance model for this kernel is a challenging task. The traditional approach of building such models by analytical modeling is difficult in practice and requires a thorough understanding of the interaction between the GPU hardware and the sparse code. In this paper, we propose to use a machine learning approach to predict the performance of the SpMV kernel using several sparse formats (COO, CSR, ELL, and HYB) on GPU. We used two popular machine learning algorithms, Support Vector Regression (SVR) and Multilayer Perceptron neural network (MLP). Our experimental results on two different GPUs (Fermi GTX 512 and Maxwell GTX 980 Ti) show that the SVR models deliver the best accuracy with average prediction error ranging between 7% and 14%.

Original languageEnglish
Title of host publicationProceedings - 22nd IEEE International Conference on Parallel and Distributed Systems, ICPADS 2016
EditorsXiaofei Liao, Robert Lovas, Xipeng Shen, Ran Zheng
PublisherIEEE Computer Society
Pages894-901
Number of pages8
ISBN (Electronic)9781509044573
DOIs
Publication statusPublished - 2 Jul 2016
Event22nd IEEE International Conference on Parallel and Distributed Systems, ICPADS 2016 - Wuhan, Hubei, China
Duration: 13 Dec 201616 Dec 2016

Publication series

NameProceedings of the International Conference on Parallel and Distributed Systems - ICPADS
Volume0
ISSN (Print)1521-9097

Conference

Conference22nd IEEE International Conference on Parallel and Distributed Systems, ICPADS 2016
Country/TerritoryChina
CityWuhan, Hubei
Period13/12/1616/12/16

Keywords

  • GPU computing
  • Multilayer Perceptron (MLP)
  • Performance modeling
  • Sparse Matrix-Vector multiplication (SpMV)
  • Support Vector Regression (SVR)

Fingerprint

Dive into the research topics of 'Machine learning approach for the predicting performance of SpMV on GPU'. Together they form a unique fingerprint.

Cite this