Compile-time partitioning of a non-strict language into sequential threads

Abstract

Presents a practical algorithm for partitioning a program into sequential threads. A thread is a sequence of instructions, possibly including branches, that can be scheduled as an indivisible unit on a von Neumann-like processor. The primary target of the proposed compilation strategy is large-scale parallel systems that rely on multithreading at the processor level to tolerate long communication latencies. As such, the algorithm incorporates a mechanism to balance the desire to maximize thread length with the desire to expose useful high-level parallelism. It can also exploit known dependency information (gathered through subscript analysis, for example). Although this paper focuses on non-strict (but not lazy) language semantics, the partitioning analysis is equally well suited to a non-strict language on a sequential machine or a strict language on a parallel multithreaded machine. >