On the working set concept for data-flow machines

Abstract

This paper discusses the concept of the working set for data-flow machines in order to establish one of the criteria for the realization of cost effective data-flow machines. The characteristics of program execution in conventional machines and data-flow machines are compared. Then, a definition of the working set for data-flow machines is proposed, based on the simultaneity of execution and the principle of locality . Several segmentation, fetch, and removal policies are described. Evaluation is made in terms of feasibility, efficiency, and performance, through computer simulations.

A pipeline bubbles reduction technique for the Monsoon dataflow architecture

“…Tokoro et uZ. [7] presented a definition of the working set for dataflow machines based on simultaneous execution and the principle of (temporal) locality, and evaluated various combinations of fetch policies and replacement policies.…”