TY - GEN
T1 - A simple performance model for multithreaded applications executing on non-uniform memory access computers
AU - Yang, R.
AU - Antony, J.
AU - Rendell, A. P.
PY - 2009
Y1 - 2009
N2 - In this work, we extend and evaluate a simple performance model to account for NUMA and bandwidth effects for single and multi-threaded calculations within the Gaussian 03 computational chemistry code on a contemporary multi-core, NUMA platform. By using the thread and memory placement APIs in Solaris, we present results for a set of calculations from which we analyze on-chip interconnect and intra-core bandwidth contention and show the importance of load-balancing between threads. The extended model predicts single threaded performance to within 1% errors and most multi-threaded experiments within 15% errors. Our results and modeling shows that accounting for bandwidth constraints within user-space code is beneficial.
AB - In this work, we extend and evaluate a simple performance model to account for NUMA and bandwidth effects for single and multi-threaded calculations within the Gaussian 03 computational chemistry code on a contemporary multi-core, NUMA platform. By using the thread and memory placement APIs in Solaris, we present results for a set of calculations from which we analyze on-chip interconnect and intra-core bandwidth contention and show the importance of load-balancing between threads. The extended model predicts single threaded performance to within 1% errors and most multi-threaded experiments within 15% errors. Our results and modeling shows that accounting for bandwidth constraints within user-space code is beneficial.
UR - http://www.scopus.com/inward/record.url?scp=70449584580&partnerID=8YFLogxK
U2 - 10.1109/HPCC.2009.39
DO - 10.1109/HPCC.2009.39
M3 - Conference contribution
SN - 9780769537382
T3 - 2009 11th IEEE International Conference on High Performance Computing and Communications, HPCC 2009
SP - 79
EP - 86
BT - 2009 11th IEEE International Conference on High Performance Computing and Communications, HPCC 2009
T2 - 11th IEEE International Conference on High Performance Computing and Communications, HPCC 2009
Y2 - 25 June 2009 through 27 June 2009
ER -