Title :
Performance model for speculative simulation using predictive optimism
Author :
Noble, B.L. ; Chamberlain, R.D.
Author_Institution :
Dept. of Electr. & Comput. Eng., Southern Illinois Univ., Edwardsville, IL, USA
Abstract :
Performance models exist that reliably describe the execution time and efficiency of discrete-event simulations executed in a synchronous iterative fashion. These performance models incorporate the effects of processor heterogeneity, other processor loads due to shared computational resources and application workload imbalance. We extend these models to include the effects of an optimistic technique known as speculative computation. This includes modeling the effects of predictive optimism, which is a technique for improving the accuracy of speculative assumptions.
Keywords :
discrete event simulation; software performance evaluation; synchronisation; application workload imbalance; discrete-event simulations; efficiency; execution time; optimistic technique; performance model; predictive optimism; processor heterogeneity; processor loads; shared computational resources; speculative assumptions accuracy; speculative computation; speculative simulation; synchronous iterative execution; Computational modeling; Concurrent computing; Costs; Discrete event simulation; Electrical capacitance tomography; Iterative algorithms; Parallel processing; Predictive models; Protocols;
Conference_Titel :
Systems Sciences, 1999. HICSS-32. Proceedings of the 32nd Annual Hawaii International Conference on
Conference_Location :
Maui, HI, USA
Print_ISBN :
0-7695-0001-3
DOI :
10.1109/HICSS.1999.773081
