Special issue on languages, compilers and environments for parallel programming Guest editors' introduction

✍ Scribed by David J. Kuck; Constantine D. Polychronopoulos

• Publisher: Elsevier Science
• Year: 1988
• Tongue: English
• Weight: 211 KB
• Volume: 5
• Category: Article
• ISSN: 0743-7315
• DOI: 10.1016/0743-7315(88)90008-1

✦ Synopsis

This issue contains papers derived from presentations at the International Conference on Supercomputing, ICS87. From the 63 papers that were published in the conference proceedings [ 11, we selected a number of software papers for additional refereeing and revision. Of these, eight were chosen for inclusion in this volume.

The hrst half of the 1980s was a period of important change in computer architecture. Many companies introduced multiple processor systems with the potential for parallel processing of single jobs. These systems ranged from supercomputers, through minisupercomputers, down to superminicomputers. In 1988 this architectural phenomenon is reaching down to the level of personal workstations.

Software to allow convenient exploitation ofthis parallelism has not enjoyed the same growth rate that the architectures themselves have. Languages and restructuring compilers for vector machines are now becoming well understood, after a decade of development. The next decade should lead to a good understanding of parallelism plus vectors. In that time frame and beyond, we expect that integrated environments for software development will also begin to mature. Applications software and operating systems are, of course, also required, but the present volume does not address these issues. We shall briefly describe the eight papers in this special issue, which deal with issues ranging from languages to programming environments.

The first paper by Arvind and K. Ekanadham addresses language issues for future parallel programming. The authors stress the importance of ease of use and the generation of efficient code. Through examples they illustrate the suitability of functional languages for parallel programming.