Coordination facilities to enhance concurrency of race free parallel algorithms
Popular styles of parallel programming are mainly characterized by simple control structure and barrier-like synchronization facilities used. However even in fairly regular cases of application programs these facilities can not provide sufficient performance. In this paper we propose to enhance concurrency of computation by means of a novel coordination model and facilities orthogonal to main body of distributed/shared memory parallel programs with static, race free structure of accesses to shared memory.
Unable to display preview. Download preview PDF.
- 1.D. Gelernter, N. Carriero, Coordination Languages and Their Significance, Commun. ACM 35, No. 2 (1992) 97–107.Google Scholar
- 2.A. E. Doroshenko, A. B. Godlevsky, Constructing Parallel Implementation with Algebraic Programming Tools // PAS'95: Proc. Aizu Int. Symp. on Parallel Algorithms/Architecture Synthesis (IEEE Comp. Sci. Press, 1995) 291–297.Google Scholar
- 3.G.R. Andrews, F.B. Schneider, Concepts and notations for concurrent programming, ACM Computing Surveys, 15, No. 1 (1983), pp. 3–43.Google Scholar
- 4.A.E. Doroshenko, Modeling synchronization and communication abstractions for dynamical parallelization, High-Performance Computing and Networking, Vienna, Austria, Apr.1997, Proc. Int. Conf., Springer Verlag, Lect. Not. in Computer Sci.,vol. 1225, 1997, pp.752–761.Google Scholar
- 5.A.E. Doroshenko, Advancing synchronization and communication techniques for distributed/shared memory parallel programs, in: PARCELLA'94: Proc. VI Int. Workshop on Parallel Processing by Cellular Arrays and Automata, Academic Verlag, Berlin, 1994, pp. 131–139.Google Scholar