A Parallelizing Compiler Based on Partial Evaluation

Abstract
This thesis demonstrates a compiler that uses partial evaluation to achieve outstandingly efficient parallel object code from very high-level source programs. The source programs are ordinary Scheme numerical programs, written abstractly, with no attempt to structure them for parallel execution. The compiler identifies and extracts parallelism completely automatically; nevertheless, it achieved by speed-ups equivalent to or better than the best observed results achieved by previous supercomputer compilers that require manual restructing of code.

This publication has 0 references indexed in Scilit: