Performance Prediction and Calibration for a Class of Multiprocessors
IEEE Transactions on Computers
Computing Maximum Task Execution Times — A Graph-BasedApproach
Real-Time Systems
Effects of communication latency, overhead, and bandwidth in a cluster architecture
Proceedings of the 24th annual international symposium on Computer architecture
Performance Modelling of Communication Networks and Computer Architectures (International Computer S
Performance Modelling of Communication Networks and Computer Architectures (International Computer S
A Callgraph-Based Search Strategy for Automated Performance Diagnosis (Distinguished Paper)
Euro-Par '00 Proceedings from the 6th International Euro-Par Conference on Parallel Processing
Exploiting fast ethernet performance in multiplatform cluster environment
Proceedings of the 2004 ACM symposium on Applied computing
Modelling asynchronous message passing in small cluster environments
International Journal of Computers and Applications
Performance-based parallel application toolkit for high-performance clusters
The Journal of Supercomputing
A dynamic and scalable performance monitoring toolkit for cluster and grid environments
International Journal of High Performance Computing and Networking
On implementation of a scalable wallet-size cluster computing system for multimedia applications
PCM'04 Proceedings of the 5th Pacific Rim conference on Advances in Multimedia Information Processing - Volume Part III
| Hi-index | 0.00 |
We present a methodology for parallel programming, along with MPI performance measurement and prediction in a class of a distributed computing environments, namely networks of workstations. Our approach is based on a two-level model where, at the top, a new parallel version of timing graph representation is used to make explicit the parallel communication and code segments of a given parallel program, while at the bottom level, analytical models are developed to represent execution behavior of parallel communications and code segments. Execution time results obtained from execution, together with problem size and number of nodes, are input to the model, which allows us to predict the performance of similar cluster computing systems with a different number of nodes. The analytical model is validated by performing experiments over a homogeneous cluster of workstations. Final results show that our approach produces accurate predictions, within 5% of actual results.