Optimal use of mixed task and data parallelism for pipelined computations
Journal of Parallel and Distributed Computing
A Pipeline-Based Approach for Scheduling Video Processing Algorithms on NOW
IEEE Transactions on Parallel and Distributed Systems
An Analytical Model for Pipeline Algorithms on Heterogeneous Clusters
Proceedings of the 9th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
PEPA nets: a structured performance modelling formalism
Performance Evaluation - Modelling techniques and tools for computer performance evaluation
PDP '05 Proceedings of the 13th Euromicro Conference on Parallel, Distributed and Network-Based Processing
Modeling master/worker applications for automatic performance tuning
Parallel Computing - Algorithmic skeletons
Modeling pipeline applications in POETRIES
Euro-Par'05 Proceedings of the 11th international Euro-Par conference on Parallel Processing
Automatic tuning of master/worker applications
Euro-Par'05 Proceedings of the 11th international Euro-Par conference on Parallel Processing
Parallelism orchestration using DoPE: the degree of parallelism executive
Proceedings of the 32nd ACM SIGPLAN conference on Programming language design and implementation
Load balancing in homogeneous pipeline based applications
Parallel Computing
MATE: toward scalable automated and dynamic performance tuning environment
PARA'10 Proceedings of the 10th international conference on Applied Parallel and Scientific Computing - Volume 2
Hi-index | 0.00 |
Parallel/distributed application development is an extremely difficult task for non-expert programmers, and support tools are therefore needed for all phases of the development cycle of this kind of applications. In particular, dynamic performance tuning tools can take advantage of the knowledge about the application's structure given by a skeleton based programming tool. This study shows the definition of a strategy for dynamically improving the performance of pipeline applications. This strategy, which has been called Dynamic Pipeline Mapping, improves the application's throughput by gathering the pipe's fastest stages and replicating its slowest ones. We have evaluated the new algorithm by experimentation and simulation, and results show that DPM leads to significant performance improvements.