A fast sparse block circulant matrix vector product

In the context of computed tomography (CT), iterative image reconstruction techniques are gaining attention because high-quality images are becoming computationally feasible. They involve the solution of large systems of equations, whose cost is dominated by the sparse matrix vector product (SpMV)....

ver descrição completa

Detalhes bibliográficos
Autores: Romero Alcalde, Eloy, Soriano Asensi, Antonio, Tomás Domínguez, Andrés Enrique, Blanquer Espert, Ignacio|||0000-0003-1692-8922
Formato: capítulo de livro
Fecha de publicación:2014
País:España
Recursos:Universitat Politècnica de València (UPV)
Repositorio:RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Idioma:inglés
OAI Identifier:oai:riunet.upv.es:10251/70427
Acesso em linha:https://riunet.upv.es/handle/10251/70427
Access Level:acceso abierto
Palavra-chave:Circulant matrix
Sparse matrix
Matrix vector product
GPU
Multi-core
Computed tomography
CIENCIAS DE LA COMPUTACION E INTELIGENCIA ARTIFICIAL
LENGUAJES Y SISTEMAS INFORMATICOS
Descrição
Resumo:In the context of computed tomography (CT), iterative image reconstruction techniques are gaining attention because high-quality images are becoming computationally feasible. They involve the solution of large systems of equations, whose cost is dominated by the sparse matrix vector product (SpMV). Our work considers the case of the sparse matrices being block circulant, which arises when taking advantage of the rotational symmetry in the tomographic system. Besides the straightforward storage saving, we exploit the circulant structure to rewrite the poor-performance SpMVs into a high-performance product between sparse and dense matrices. This paper describes the implementations developed for multi-core CPUs and GPUs, and presents experimental results with typical CT matrices. The presented approach is up to ten times faster than without exploiting the circulant structure.