Parallelism control scheme in a dataflow architecture
We propose execution control scheme to realize an efficient multi-processing environment in dataflow architecture. With a unified software and hardware mechanism, we can detect states of processes at run time, which are treated as control information. We control the process level parallelism according to the hardware capacity, and virtualize high speed memory in dynamic data-driven computation where context changes in every instruction execution. We apply this scheme to the Datarol architecture, an optimized version of dynamic dataflow architecture, and evaluate them through software simulation.
KeywordsMemory Management Register File Memory Unit Unify Software Terminal Operation
Unable to display preview. Download preview PDF.
- M.Amamiya, and R.Taniguchi: “Datarol: A Massively Parallel Architecture for Functional Language”, Proc. IEEE 2nd SPDP, pp.726–735 (Dec. 1990).Google Scholar
- V.G.Grafe and J.E. Hoch: “The Epsilon-2 Multi-Processor System”, Journal of Parallel and Distributed Computing, 10, pp.309–318 (Oct. 1990).Google Scholar
- G.M.Papadopoulos and E.C.Culler: “Monsoon: an Explicit Token-Store Architecture”, Proc. 17th ISCA, pp.82–91 (June 1990).Google Scholar
- C.A.Ruggiero and J.Sargeant: “Control of Parallelism in the Manchester Dataflow Machine”, Proc. FPCA (Lecture Notes in Computer Science 274, Springer-Verlag), pp.1–15, (Sep. 1987).Google Scholar
- S.Sakai, Y.Yamaguchi, K.Hiraki, Y.Kodama and T.Yuba: “An Architecture of a Dataflow Single Chip Processor”, Proc. 16th ISCA, pp.46–53 (May 1989).Google Scholar
- T.Tachibana, R.Taniguchi and Amamiya,M.: “Compiling Method of Functional Programming Language Valid by Data Flow Analysis-Extraction of Datarol Program-”, Journal of Information Processing(in Japanese), Vol.30, No.12, pp.1628–1638 (Dec. 1989).Google Scholar
- M.Takesue: “A Unified Resource Management and Execution Control Mechanism for Data Flow Machines”, Proc. 14th ISCA, pp.90–97 (Jun. 1987).Google Scholar