Performance and Energy Analysis of OpenMP Runtime Systems with Dense Linear Algebra Algorithms

In this paper, we analyse performance and energy consumption of four OpenMP runtime systems over a NUMA platform. We present an experimental study to characterize OpenMP runtime systems on the three main kernels in dense linear algebra algorithms (Cholesky, LU and QR) in terms of performance and ene...

Full description

Saved in:
Bibliographic Details
Published in2017 International Symposium on Computer Architecture and High Performance Computing Workshops (SBAC-PADW) pp. 7 - 12
Main Authors Lima, Joao V. F., Rais, Issam, Lefevre, Laurent, Gautier, Thierry
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2017
Subjects
Benchmark testing
Energy consumption
Kernel
Linear algebra
NUMA
OpenMP
Performance Energy
Program processors
Runtime
Task parallelism
Online AccessGet full text

Cover

More Information
Summary:In this paper, we analyse performance and energy consumption of four OpenMP runtime systems over a NUMA platform. We present an experimental study to characterize OpenMP runtime systems on the three main kernels in dense linear algebra algorithms (Cholesky, LU and QR) in terms of performance and energy consumption. Our experimental results suggest that OpenMP runtime systems can be considered as a new energy leverage. For instance, a LU factorization with concurrent write extension from libKOMP achieved up to 1.75 of performance gain and 1.56 of energy decrease.
DOI:10.1109/SBAC-PADW.2017.10