Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU Architecture

Sparse matrix–vector multiplication (SpMV) plays a significant role in the computational costs of many scientific applications such as 2D/3D robotics, power network problems, and computer vision. Numerous implementations using different sparse matrix formats have been introduced to optimize this ker...

Full description

Saved in:
Bibliographic Details
Main Authors: Muhammad Ahmad, Sardar Usman, Ameer Hamza, Muhammad Muzamil, Ildar Batyrshin
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Information
Subjects:
Online Access:https://www.mdpi.com/2078-2489/16/7/553
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849406666057973760
author Muhammad Ahmad
Sardar Usman
Ameer Hamza
Muhammad Muzamil
Ildar Batyrshin
author_facet Muhammad Ahmad
Sardar Usman
Ameer Hamza
Muhammad Muzamil
Ildar Batyrshin
author_sort Muhammad Ahmad
collection DOAJ
description Sparse matrix–vector multiplication (SpMV) plays a significant role in the computational costs of many scientific applications such as 2D/3D robotics, power network problems, and computer vision. Numerous implementations using different sparse matrix formats have been introduced to optimize this kernel on CPUs and GPUs. However, due to the sparsity patterns of matrices and the diverse configurations of hardware, accurately modeling the performance of SpMV remains a complex challenge. SpMV computation is often a time-consuming process because of its sparse matrix structure. To address this, we propose a machine learning-based tool, namely Elegante+, that predicts optimal scheduling policies by analyzing matrix structures. This approach eliminates the need for repetitive trial and error, minimizes errors, and finds the best solution of the SpMV kernel, which enables users to make informed decisions about scheduling policies that maximize computational efficiency. For this purpose, we collected 1000+ sparse matrices from the SuiteSparse matrix market collection and converted them into the compressed sparse row (CSR) format, and SpMV computation was performed by extracting 14 key sparse matrix features. After creating a comprehensive dataset, we trained various machine learning models to predict the optimal scheduling policy, significantly enhancing the computational efficiency and reducing the overhead in high-performance computing environments. Our proposed tool, Elegante+ (XGB with all SpMV features), achieved the highest cross-validation score of 79% and performed five times faster than the default scheduling policy during SpMV in a high-performance computing (HPC) environment.
format Article
id doaj-art-e307b7abf84d463582a2005c0681a4aa
institution Kabale University
issn 2078-2489
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Information
spelling doaj-art-e307b7abf84d463582a2005c0681a4aa2025-08-20T03:36:18ZengMDPI AGInformation2078-24892025-06-0116755310.3390/info16070553Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU ArchitectureMuhammad Ahmad0Sardar Usman1Ameer Hamza2Muhammad Muzamil3Ildar Batyrshin4Centro de Investigación en Computación, Instituto Politécnico Nacional (CIC-IPN), Mexico City 07738, MexicoSchool of Informatics and Robotics, Institute of Arts and Culture, Lahore 54000, PakistanDepartment of Computer Science and Software Engineering, The Islamia University of Bahawapur, Bahawalpur 63100, PakistanDepartment of Computer Science and Software Engineering, The Islamia University of Bahawapur, Bahawalpur 63100, PakistanCentro de Investigación en Computación, Instituto Politécnico Nacional (CIC-IPN), Mexico City 07738, MexicoSparse matrix–vector multiplication (SpMV) plays a significant role in the computational costs of many scientific applications such as 2D/3D robotics, power network problems, and computer vision. Numerous implementations using different sparse matrix formats have been introduced to optimize this kernel on CPUs and GPUs. However, due to the sparsity patterns of matrices and the diverse configurations of hardware, accurately modeling the performance of SpMV remains a complex challenge. SpMV computation is often a time-consuming process because of its sparse matrix structure. To address this, we propose a machine learning-based tool, namely Elegante+, that predicts optimal scheduling policies by analyzing matrix structures. This approach eliminates the need for repetitive trial and error, minimizes errors, and finds the best solution of the SpMV kernel, which enables users to make informed decisions about scheduling policies that maximize computational efficiency. For this purpose, we collected 1000+ sparse matrices from the SuiteSparse matrix market collection and converted them into the compressed sparse row (CSR) format, and SpMV computation was performed by extracting 14 key sparse matrix features. After creating a comprehensive dataset, we trained various machine learning models to predict the optimal scheduling policy, significantly enhancing the computational efficiency and reducing the overhead in high-performance computing environments. Our proposed tool, Elegante+ (XGB with all SpMV features), achieved the highest cross-validation score of 79% and performed five times faster than the default scheduling policy during SpMV in a high-performance computing (HPC) environment.https://www.mdpi.com/2078-2489/16/7/553SpMVmachine learninghigh-performance computingSVMCSRsparse matrix
spellingShingle Muhammad Ahmad
Sardar Usman
Ameer Hamza
Muhammad Muzamil
Ildar Batyrshin
Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU Architecture
Information
SpMV
machine learning
high-performance computing
SVM
CSR
sparse matrix
title Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU Architecture
title_full Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU Architecture
title_fullStr Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU Architecture
title_full_unstemmed Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU Architecture
title_short Elegante+: A Machine Learning-Based Optimization Framework for Sparse Matrix–Vector Computations on the CPU Architecture
title_sort elegante a machine learning based optimization framework for sparse matrix vector computations on the cpu architecture
topic SpMV
machine learning
high-performance computing
SVM
CSR
sparse matrix
url https://www.mdpi.com/2078-2489/16/7/553
work_keys_str_mv AT muhammadahmad eleganteamachinelearningbasedoptimizationframeworkforsparsematrixvectorcomputationsonthecpuarchitecture
AT sardarusman eleganteamachinelearningbasedoptimizationframeworkforsparsematrixvectorcomputationsonthecpuarchitecture
AT ameerhamza eleganteamachinelearningbasedoptimizationframeworkforsparsematrixvectorcomputationsonthecpuarchitecture
AT muhammadmuzamil eleganteamachinelearningbasedoptimizationframeworkforsparsematrixvectorcomputationsonthecpuarchitecture
AT ildarbatyrshin eleganteamachinelearningbasedoptimizationframeworkforsparsematrixvectorcomputationsonthecpuarchitecture