Title :
On the Cost of a General GPU Framework: The Strange Case of CUDA 4.0 vs. CUDA 5.0
Author :
Wezowicz, M. ; Taufer, Michela
Author_Institution :
Univ. of Delaware, Newark, DE, USA
Abstract :
The first release of CUDA was in 2007. Since then, we have experienced frequent new releases. CUDA reached its max performance with CUDA 4.0. Since its release, NVIDIA has started a re-design of the CUDA framework driven by software engineering prospective, i.e., the search for a general, multi-layer framework whose compiler back-end is unified with OpenCL. This can have significant impact on both maintenance costs and cross-platform portability. The software engineering community applauded this direction. However, our poster indicates that the new direction comes at a high cost in performance. Supported by the rigorous performance analysis of a MD code and its optimizations, we want to provocatively raise the question: Is the HPC community, that has been benefiting the most from CUDA and GPU programming, ready to pay this cost?
Keywords :
graphics processing units; parallel architectures; program compilers; software engineering; CUDA 4.0; CUDA 5.0; GPU programming; HPC community; MD code; NVIDIA; OpenCL; compiler back-end; cross-platform portability; general GPU framework; maintenance costs; multilayer framework; software engineering prospective; CUDA; GPGPU;
Conference_Titel :
High Performance Computing, Networking, Storage and Analysis (SCC), 2012 SC Companion:
Conference_Location :
Salt Lake City, UT
Print_ISBN :
978-1-4673-6218-4
DOI :
10.1109/SC.Companion.2012.310
