TY - GEN
T1 - Developing performance-portable molecular dynamics kernels in OpenCL
AU - Pennycook, S. J.
AU - Jarvis, S. A.
PY - 2012
Y1 - 2012
N2 - This paper investigates the development of a molecular dynamics code that is highly portable between architectures. Using OpenCL, we develop an implementation of Sandia's miniMD benchmark that achieves good levels of performance across a wide range of hardware: CPUs, discrete GPUs and integrated GPUs. We demonstrate that the performance bottlenecks of miniMD's short-range force calculation kernel are the same across these architectures, and detail a number of platform-Agnostic optimisations that improve its performance by at least 2x on all hardware considered. Our complete code is shown to be 1.7x faster than the original miniMD, and at most 2x slower than implementations individually hand-tuned for a specific architecture.
AB - This paper investigates the development of a molecular dynamics code that is highly portable between architectures. Using OpenCL, we develop an implementation of Sandia's miniMD benchmark that achieves good levels of performance across a wide range of hardware: CPUs, discrete GPUs and integrated GPUs. We demonstrate that the performance bottlenecks of miniMD's short-range force calculation kernel are the same across these architectures, and detail a number of platform-Agnostic optimisations that improve its performance by at least 2x on all hardware considered. Our complete code is shown to be 1.7x faster than the original miniMD, and at most 2x slower than implementations individually hand-tuned for a specific architecture.
KW - accelerator architectures
KW - high performance computing
KW - parallel programming
KW - performance analysis
KW - scientific computing
UR - http://www.scopus.com/inward/record.url?scp=84876546455&partnerID=8YFLogxK
U2 - 10.1109/SC.Companion.2012.58
DO - 10.1109/SC.Companion.2012.58
M3 - Conference contribution
AN - SCOPUS:84876546455
SN - 9780769549569
T3 - Proceedings - 2012 SC Companion: High Performance Computing, Networking Storage and Analysis, SCC 2012
SP - 386
EP - 395
BT - Proceedings - 2012 SC Companion
T2 - 2012 SC Companion: High Performance Computing, Networking Storage and Analysis, SCC 2012
Y2 - 10 November 2012 through 16 November 2012
ER -