Towards a Cloud Friendly Adaptive Parallel Runtime for HPC Applications

Cloud computing is increasingly being explored as a cost effective alternative and addition to supercomputers for some High Performance Computing (HPC) applications. However, poor interconnect performance; commodity operating systems, dynamic environment and interference by other virtual machines are some of the factors which prevent efficient execution of HPC applications in cloud. Through this paper, the authors propose to leverage a message driven adaptive runtime system to develop techniques to reduce the gap between application performance on cloud and HPC optimized systems and utilize cloud specific features such as variable pricing.

Provided by: University of Idaho Topic: Data Centers Date Added: Mar 2012 Format: PDF

Find By Topic