A Feasibility Study on Porting the Community Land Model onto Accelerators Using Openacc
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2014, Vol 5, Issue 12
Abstract
As environmental models (such as Accelerated Climate Model for Energy (ACME), Parallel Reactive Flow and Transport Model (PFLOTRAN), Arctic Terrestrial Simulator (ATS), etc.) became more and more complicated, we are facing enormous challenges regarding to porting those applications onto hybrid computing architecture. OpenACC emerges as a very promising technology, therefore, we have conducted a feasibility analysis on porting the Community Land Model (CLM), a terrestrial ecosystem model within the Community Earth System Models (CESM)). Specifically, we used automatic function testing platform to extract a small computing kernel out of CLM, then we apply this kernel into the actually CLM dataflow procedure, and investigate the strategy of data parallelization and the benefit of data movement provided by current implementation of OpenACC. Even it is a non-intensive kernel, on a single 16-core computing node, the performance (based on the actual computation time using one GPU) of OpenACC implementation is 2.3 time faster than that of OpenMP implementation using single OpenMP thread, but it is 2.8 times slower than the performance of OpenMP implementation using 16 threads. On multiple nodes, MPI_OpenACC implementation demonstrated very good scalability on up to 128 GPUs on 128 computing nodes. This study also provides useful information for us to look into the potential benefits of “deep copy” capability and “routine” feature of OpenACC standards. We believe that our experience on the environmental model, CLM, can be beneficial to many other scientific research programs who are interested to porting their large scale scientific code using OpenACC onto high-end computers, empowered by hybrid computing architecture.
Authors and Affiliations
D. Wang, W. Wu, F. Winkler, O. Hernandez
Keywords
Related Articles
- EP ID EP100184
- DOI 10.14569/IJACSA.2014.051203
- Views 97
- Downloads 0
Energy-Efficient, Noise-Tolerant CMOS Domino VLSI Circuits in VDSM Technology
Compared to static CMOS logic, dynamic logic offers good performance. Wide fan-in dynamic logic such as domino is often used in performance critical paths, to achieve high speeds where static CMOS fails to meet per...
Development of Eye-Blink and Face Corpora for Research in Human Computer Interaction
A major requirement in face recognition research and coded voluntary eye-blink based sign language communication research is a robust face and eye-blink image corpora. The effectiveness, confidence level, and acceptabili...
Optimized Order of Software Testing Techniques in Agile Process – A Systematic Approach
The designing, development of a software product needs lot of efforts whereas software testing is also a very challenging task but it is equally mandatory activity in order to ensure the quality of the product before shi...
Semantic Conflicts Reconciliation as a Viable Solution for Semantic Heterogeneity Problems
Achieving semantic interoperability is a current challenge in the field of data integration in order to bridge semantic conflicts occurring when the participating sources and receivers use different or implicit data assu...
Optimized Voting Scheme for Efficient Vanishing Point Detection in General Road Images
Next generation automobile industries are aiming for development of vision-based driver assistance system and driver-less vehicle system. In the context of this application, a major challenge lies in the identification o...