skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: On the Impact of Execution Models: A Case Study in Computational Chemistry

Abstract

Efficient utilization of high-performance computing (HPC) platforms is an important and complex problem. Execution models, abstract descriptions of the dynamic runtime behavior of the execution stack, have significant impact on the utilization of HPC systems. Using a computational chemistry kernel as a case study and a wide variety of execution models combined with load balancing techniques, we explore the impact of execution models on the utilization of an HPC system. We demonstrate a 50 percent improvement in performance by using work stealing relative to a more traditional static scheduling approach. We also use a novel semi-matching technique for load balancing that has comparable performance to a traditional hypergraph-based partitioning implementation, which is computationally expensive. Using this study, we found that execution model design choices and assumptions can limit critical optimizations such as global, dynamic load balancing and finding the correct balance between available work units and different system and runtime overheads. With the emergence of multi- and many-core architectures and the consequent growth in the complexity of HPC platforms, we believe that these lessons will be beneficial to researchers tuning diverse applications on modern HPC platforms, especially on emerging dynamic platforms with energy-induced performance variability.

Authors:
; ; ; ; ;
Publication Date:
Research Org.:
Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1236313
Report Number(s):
PNNL-SA-108382
KJ0402000
DOE Contract Number:  
AC05-76RL01830
Resource Type:
Conference
Resource Relation:
Conference: Joint International Workshop on High-level Parallel Programming Models and Supportive Environments (HIPS) and Large-Scale Parallel Processing (LSPP), held in conjunction with the 29th IEEE International Parallel & Distributed Processing Symposium Workshop (IPDPSW 2015), May 25-29, 2015, Hyderabad, India, 255-264
Country of Publication:
United States
Language:
English
Subject:
Execution models, computational chemistry, MPI and CSP

Citation Formats

Chavarría-Miranda, Daniel, Halappanavar, Mahantesh, Krishnamoorthy, Sriram, Manzano Franco, Joseph B., Vishnu, Abhinav, and Hoisie, Adolfy. On the Impact of Execution Models: A Case Study in Computational Chemistry. United States: N. p., 2015. Web. doi:10.1109/IPDPSW.2015.111.
Chavarría-Miranda, Daniel, Halappanavar, Mahantesh, Krishnamoorthy, Sriram, Manzano Franco, Joseph B., Vishnu, Abhinav, & Hoisie, Adolfy. On the Impact of Execution Models: A Case Study in Computational Chemistry. United States. doi:10.1109/IPDPSW.2015.111.
Chavarría-Miranda, Daniel, Halappanavar, Mahantesh, Krishnamoorthy, Sriram, Manzano Franco, Joseph B., Vishnu, Abhinav, and Hoisie, Adolfy. Mon . "On the Impact of Execution Models: A Case Study in Computational Chemistry". United States. doi:10.1109/IPDPSW.2015.111.
@article{osti_1236313,
title = {On the Impact of Execution Models: A Case Study in Computational Chemistry},
author = {Chavarría-Miranda, Daniel and Halappanavar, Mahantesh and Krishnamoorthy, Sriram and Manzano Franco, Joseph B. and Vishnu, Abhinav and Hoisie, Adolfy},
abstractNote = {Efficient utilization of high-performance computing (HPC) platforms is an important and complex problem. Execution models, abstract descriptions of the dynamic runtime behavior of the execution stack, have significant impact on the utilization of HPC systems. Using a computational chemistry kernel as a case study and a wide variety of execution models combined with load balancing techniques, we explore the impact of execution models on the utilization of an HPC system. We demonstrate a 50 percent improvement in performance by using work stealing relative to a more traditional static scheduling approach. We also use a novel semi-matching technique for load balancing that has comparable performance to a traditional hypergraph-based partitioning implementation, which is computationally expensive. Using this study, we found that execution model design choices and assumptions can limit critical optimizations such as global, dynamic load balancing and finding the correct balance between available work units and different system and runtime overheads. With the emergence of multi- and many-core architectures and the consequent growth in the complexity of HPC platforms, we believe that these lessons will be beneficial to researchers tuning diverse applications on modern HPC platforms, especially on emerging dynamic platforms with energy-induced performance variability.},
doi = {10.1109/IPDPSW.2015.111},
journal = {},
number = ,
volume = ,
place = {United States},
year = {2015},
month = {5}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share: