Part I: Special Issue on Parallel Architectures and Compilation Techniques

Abstract

Welcome to part one of this special issue on Parallel Architectures and Compilation Techniques. The papers you will find here are a compendium of some (by no means the only) of the most relevant and strongest papers which were presented at the Conference on Parallel Architectures and Compilation Techniques (PACT '95). A word about PACT '95: it met in Cyprus, in June 1995, and it was the third in a series of meetings which were founded with a unique goal in mind: to provide a forum for the interaction of researchers working at the precise boundary between architecture and compilers for parallel machines. Now back to this special issue: the papers it contains are a good sample of the topics covered by the conference as they range from analytical modeling to parallel compilation techniques, to the design of parallel architectures. The first paper, "Using predicated execution to improve the performance of a dynamically scheduled machine with speculative execution" by P-Y. Chang, E. Hao, Y. Patt, and P. Chang, recognizes that two mechanisms for handling conditional branches, namely speculative execution and predicated execution, have very different characteristics and uses, and shows how a combination of the two can be used to advantage in wide-issue, deeply pipelined processors. In the second paper, "A mean value analysis multiprocessor model incorporating superscalar processors and latency-tolerating techniques," D. Albonesi and 1. Koren have demonstrated a Mean Value Analysis model to evaluate the tradeoffs involved in the design of multiprocessors whose nodes are individual superscalar processors. How to generate code to enumerate the integer points of a polyhedron is the topic of the third paper, "A simple algorithm for the generation of efficient loop structures," by M, Cosnard and M. Loi. It is a known important

DOI: 10.1007/BF03356748

Cite this paper

@article{Gaudiot2016PartIS, title={Part I: Special Issue on Parallel Architectures and Compilation Techniques}, author={Jean-Luc Gaudiot}, journal={International Journal of Parallel Programming}, year={2016}, volume={24}, pages={207-208} }