Experiment of Multithreading Symbolic and Algebraic Computations with OpenMP

This paper describes the current status of a project for multithreading algebraic computations, which aims at the utilization of today’s high-spec PCs with hyperthreading or dual-core technologies. Our effort is done by applying OpenXM with minimal cost of development, and includes memory management...

Full description

Saved in:
Bibliographic Details
Published inLecture notes in computer science pp. 426 - 437
Main Author Murao, Hirokazu
Format Book Chapter Conference Proceeding
LanguageEnglish
Published Berlin, Heidelberg Springer Berlin Heidelberg 2006
Springer
SeriesLecture Notes in Computer Science
Subjects
Algebraic Computation
Algorithmics. Computability. Computer arithmetics
Applied sciences
Computer Algebra System
Computer science; control theory; systems
Exact sciences and technology
Garbage Collection
Memory Management
Polynomial Multiplication
Theoretical computing
Mathematical function
Multiprocessor
Multithread
Program library
Storage management
Development cost
Symbolic computation
Shared memory
Online AccessGet full text

Cover

More Information
Summary:This paper describes the current status of a project for multithreading algebraic computations, which aims at the utilization of today’s high-spec PCs with hyperthreading or dual-core technologies. Our effort is done by applying OpenXM with minimal cost of development, and includes memory management in multithreaded environment. Our empirical results show that the performance gain can be attained in numeric cases and in some cases of purely symbolic computations.
ISBN:9783540380849
3540380841
ISSN:0302-9743
1611-3349
DOI:10.1007/11832225_44