Performance and Reliability Analysis Using Directed Acyclic Graphs
IEEE Transactions on Software Engineering
A bridging model for parallel computation
Communications of the ACM
A cost calculus for parallel functional programming
Journal of Parallel and Distributed Computing
Scheduling UET-UCT series-parallel graphs on two processors
Theoretical Computer Science
The importance of synchronization structure in parallel program optimization
ICS '97 Proceedings of the 11th international conference on Supercomputing
Models and languages for parallel computation
ACM Computing Surveys (CSUR)
Thread fork/join techniques for multi-level parallelism exploitation in NUMA multiprocessors
ICS '99 Proceedings of the 13th international conference on Supercomputing
Functional Skeletons for Parallel Coordination
Euro-Par '95 Proceedings of the First International Euro-Par Conference on Parallel Processing
Scheduling multithreaded computations by work stealing
SFCS '94 Proceedings of the 35th Annual Symposium on Foundations of Computer Science
| Hi-index | 0.00 |
A number of interesting properties for scheduling and/or cost estimation arise when using parallel programming models that restrict the topology of a program's task graph to an SP (series-parallel) form. A critical question however, is to what extent the ability to exploit parallelism is compromised when only SP coordination structures are allowed. This paper presents new application parameters which are key factors to predict this loss of parallelism at both language modeling and program execution levels, for shared-memory architectures. Our results indicate that a wide range of parallel computations can be expressed using a structured coordination model with a loss of parallelism that is small and predictable.