Sciweavers

ICPP
2007
IEEE

COBRA: An Adaptive Runtime Binary Optimization Framework for Multithreaded Applications

13 years 10 months ago
COBRA: An Adaptive Runtime Binary Optimization Framework for Multithreaded Applications
This paper presents COBRA (Continuous Binary ReAdaptation), a runtime binary optimization framework, for multithreaded applications. It is currently implemented on Itanium 2 based SMP and cc-NUMA systems. Using OpenMP NAS parallel benchmark, we show how COBRA can adaptively choose appropriate optimizations according to observed changing runtime program behavior. Coherent cache misses caused by true/false data sharing often limit the scalability of multithreaded applications. This paper shows that COBRA can significantly improve the performance of some applications parallelized with OpenMP, by reducing the aggressiveness of data prefetching and by using exclusive hints for prefetch instructions. For example, we show that COBRA can improve the performance of OpenMP NAS parallel benchmarks up to 68%, with an average of 17.5% on the SGI Altix cc-NUMA system.
Jinpyo Kim, Wei-Chung Hsu, Pen-Chung Yew
Added 03 Jun 2010
Updated 03 Jun 2010
Type Conference
Year 2007
Where ICPP
Authors Jinpyo Kim, Wei-Chung Hsu, Pen-Chung Yew
Comments (0)