Parallel Algorithms / Architecture Synthesis, AIZU International Symposium on (1995)
Aizu-Wakamatsu, Fukushima, Japan
Mar. 15, 1995 to Mar. 17, 1995
W. Lowe , Fakultat fur Inf., Karlsruhe Univ., Germany
W. Zimmermann , Fakultat fur Inf., Karlsruhe Univ., Germany
Currently, many parallel algorithms are defined for shared memory architectures. The preferred machine model is the PRAM, but this model does not take into account properties of existing architectures that have a distributed memory and an asynchronous execution model. A transformation of PRAM programs into distributed, asynchronous ones is known. In order to produce not only correct but also efficient code the tasks have to be clustered. We introduce a parallel algorithm producing an optimal clustering for coarse grained task graphs with respect to the execution time on an asynchronous distributed random access machine, the A-DRAM. This machine model assumes distributed memory, asynchronous execution of tasks, computation costs, and communication delay.
parallel algorithms; distributed memory systems; random-access storage; graph theory; storage management; optimal clusterings; task graphs; parallel algorithms; shared memory architectures; PRAM; distributed memory; asynchronous execution model; optimal clustering; coarse grained task graphs; asynchronous distributed random access machine; A-DRAM; distributed memory asynchronous task execution
W. Lowe and W. Zimmermann, "On finding optimal clusterings of task graphs," Parallel Algorithms / Architecture Synthesis, AIZU International Symposium on(PAS), Aizu-Wakamatsu, Fukushima, Japan, 1995, pp. 241.