Title :
Parallel history sensitive computations in dataflow architecture
Author :
Park, Jin H. ; George, K.M.
Author_Institution :
Dept. of Comput. Sci., Oklahoma State Univ., Stillwater, OK, USA
Abstract :
This paper addresses the history sensitive computation problem in dataflow architecture. Current solutions use memory which is going a step backwards to Von Neumann model of computers. Furthermore, these solutions are not suitable for exploiting parallelism. In this paper, a new memoryless solution is reviewed and approaches based on forwarding, pipelining, and loop unfolding to maximize its parallel processing are presented. Forwarding mechanisms are presented for both static and dynamic dataflow environments
Keywords :
data flow computing; parallel architectures; dataflow architecture; forwarding; history sensitive computations; loop unfolding; memoryless solution; parallelism; pipelining; Automata; Centralized control; Computational modeling; Computer aided instruction; Computer architecture; Computer science; Concurrent computing; High performance computing; History; Parallel processing;
Conference_Titel :
Algorithms & Architectures for Parallel Processing, 1996. ICAPP 96. 1996 IEEE Second International Conference on
Print_ISBN :
0-7803-3529-5
DOI :
10.1109/ICAPP.1996.562918
