Crystallizing FORTRAN
Abstract
The parallelization techniques embodied in source to source tools (that are customarily applied to FORTRAN programs) and in compilers for high level functional languages (such as Crystal) can and should be combined under a common framework. traditionally, the former tools target shared memory machines, whereas the latter - distributed memory (or data parallel) machines. One effect of the combination is the ability to run sequential FORTRAN programs on a whole new class of machines; it also provides a good way to separate between different parallelization concerns and deal with them at the appropriate level. We describe the design and an implementation of a prototypical system that enables such a combination. The design is modular and extendible; its implementation entails, among other challenges, bridging the gap between imperative and functional program semantics. We have developed several techniques that solve these and other fundamental problems, and present them with this concrete framework. Experimental results, as achieved by the prototype and as presented hereby, are most encouraging and substantiate the usefulness of this approach.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 26, 1991
- Accession Number
- ADA250162
Entities
People
- Dong Yuan Chen
- Marina C. Chen
- Ron Y. Pinter
- Shlomit S. Pinter
Organizations
- Yale University