Simulation of Parallel Applications on Large-scale Distributed Systems
MetadataPokaż pełny rekord
This chapter has a form of a review article in the field of simulating High Performance Computing systems. We justify the need for a new versatile simulator considering heterogeneity, energy efficiency and reliability of HPC systems. We sketch the problems that need to be solved by such simulator and rationalize using discrete-event simulation for this purpose. Based on a review of existing discrete-event HPC simulation solutions we propose a flexible application execution model. Finally, we outline new directions of developing the methods to generate the traces of distributed application executions.