DocumentCode :
244762
Title :
Performance and energy efficiency in material science simulation on heterogeneous architectures
Author :
Pascolo, Eric ; Affinito, Fabio ; Cavazzoni, Carlo
Author_Institution :
DIFA, Univ. of Bologna, Bologna, Italy
fYear :
2014
fDate :
21-25 July 2014
Firstpage :
927
Lastpage :
932
Abstract :
In HPC applications, the energy efficiency is becoming more and more important, due to architectural constraints. It is therefore of primary interest to measure and evaluate the energy efficiency of current architectures using typical HPC workloads. One of the most used and appreciated codes publicly available for computational material science simulation, and largely used in many high end HPC system is Quantum ESPRESSO [1]. In this work we present the results of a set of benchmarks of Quantum ESPRESSO on the heterogeneous EURORA Supercomputer, featuring both NVIDIA K20 GPU and Xeon PHI co-processors, and measuring the energy to solution, thanks to an innovative monitoring system recently deployed on EURORA.
Keywords :
energy conservation; graphics processing units; materials science computing; multiprocessing systems; parallel processing; power aware computing; HPC applications; HPC workload; NVIDIA K20 GPU; Quantum ESPRESSO system; Xeon PHI coprocessors; energy efficiency; graphics processing unit; heterogeneous EURORA supercomputer; heterogeneous architectures; high performance computing; material science simulation; Benchmark testing; Computational modeling; Computer architecture; Coprocessors; Graphics processing units; Kernel; Materials; Energy efficiency; accelerators; benchmark; heterogeneous architectures; material science; parallel;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
High Performance Computing & Simulation (HPCS), 2014 International Conference on
Conference_Location :
Bologna
Print_ISBN :
978-1-4799-5312-7
Type :
conf
DOI :
10.1109/HPCSim.2014.6903788
Filename :
6903788
Link To Document :
بازگشت