From METCALF@crnvma.cern.ch Mon Jun 29 15:32:15 1992 Received: from danpost2.uni-c.dk by dkuug.dk via EUnet with SMTP (5.64+/8+bit/IDA-1.2.8) id AA11373; Mon, 29 Jun 92 15:32:15 +0200 Received: from vm.uni-c.dk by danpost2.uni-c.dk (5.65/1.34) id AA04736; Mon, 29 Jun 92 15:30:34 +0200 Resent-Message-Id: <9206291330.AA04736@danpost2.uni-c.dk> Received: from vm.uni-c.dk by vm.uni-c.dk (IBM VM SMTP V2R1) with BSMTP id 9701; Mon, 29 Jun 92 15:32:36 DNT Received: from CERNVM.cern.ch by vm.uni-c.dk (Mailer R2.07) with BSMTP id 2231; Mon, 29 Jun 92 15:32:34 DNT Received: from CERNVM.CERN.CH (METCALF) by CERNVM.cern.ch (Mailer R2.08) with BSMTP id 7229; Mon, 29 Jun 92 15:31:58 SET Resent-Date: Mon, 29 Jun 92 15:27:48 SET Resent-From: Michael Metcalf Resent-To: sc22wg5@dkuug.dk Received: from CERNVM by CERNVM.cern.ch (Mailer R2.08) with BSMTP id 2547; Fri, 26 Jun 92 13:40:32 SET Received: from dxmint.cern.ch by CERNVM.CERN.CH (IBM VM SMTP V2R2) with TCP; Fri, 26 Jun 92 13:40:31 SET Received: by dxmint.cern.ch (dxcern) (5.57/3.14) id AA28831; Fri, 26 Jun 92 13:40:19 +0200 Received: from slim.gmd.de (slim) by gmdzi.gmd.de with SMTP id AA18075 (5.65c/IDA-1.4.4 for ); Fri, 26 Jun 1992 13:21:51 +0200 Received: by slim.gmd.de id AA23538 (5.65c/IDA-1.4.3 for hpf-europe@gmdzi.gmd.de); Fri, 26 Jun 1992 13:22:25 +0200 Date: 26 Jun 92 12:11 From: Brian Wylie Sender: bjnw@castle.edinburgh.ac.uk To: hpf-europe@gmdzi.gmd.de Cc: MGN@edinburgh.ac.uk, LJC@edinburgh.ac.uk Message-Id: <9206261211.aa20603(a)castle.ed.ac.uk> Subject: HPF Perspective Technical Note X-Charset: ASCII X-Char-Esc: 29 This is of more than passing interest to f90 fans. Mike Metcalf P.S. Technical note: I had to type cd pub/tn/92 get tn9205.ps.Z to get ftp to work. P.P.S. What is the current status of the DoD's threat to boycott f90? ----------------------------Original message---------------------------- RFC-822-HEADERS: Organisation: Edinburgh Parallel Computing Centre (EPCC) ================== In an attempt to clarify the HPF developments in progress, primarily for the benefit of EPCC staff and associates, two colleagues and myself prepared a Technical Note outlining our perspective of the history, current state, and some likely extensions of the current HPF proposal. This note may be a useful introduction to HPF for those who haven't been able to attend the meetings or keep track of the mailing lists, as an interrim measure prior to the release of an official HPF document (which I believe is currently being drafted by Chuck Koelbel). I would therefore like to take advantage of some of the currently free bandwidth on hpf-europe to publicise it's availability via anonymous ftp. [The abstract is included at the end of this message.] Connect to epcc.ed.ac.uk (129.215.56.29) using the "anonymous" login and your email address as password, and ftp the following compressed PostScript file (remembering to set binary file transfer mode): epcc.ed.ac.uk:pub/tn/92/tn9205.ps.Z While this document is certainly incomplete (not least in the fact that it only covers the period up to the end of May, and there have been both European and US meetings since then) and possibly mistaken in certain respects, we hope that others might find it useful, and would value any comments regarding its content. Slainte, Brian J N Wylie | EPCC, University of Edinburgh, JCMB, King's Bldgs, Edinburgh, | EH9 3JZ, Scotland [Tel: +44 31 650-5962, Fax: +44 31 650-6555] ================================================================================ EPCC-TN92-05 High Performance Fortran: A Perspective Brian J N Wylie Michael G Norman Lyndon J Clarke May 1992 High Performance Fortran (HPF) promises an attractive abstraction, with high-performance potential, for Fortran programs on parallel (and massively-parallel) computers. This report is intended to provide background information on the HPF initiative, the technical aspects thereof, as well as our impressions of why this activity is progressing with the apparent enthusiasm of a number of parallel computer vendors and other interested parties. It is contended that the array-parallel operations of Fortran-90, coupled with additional directives providing information about appropriate decompositions and distributions, can enable compilers with sophisticated dependence analysis and knowledge of the MIMD parallelisation techniques to produce native message-passing code for efficient and effective utilisation of distributed-memory parallel computers. In addition to providing a perspective of the developing HPF situation in the United States, European activities and involvement are also addressed. It seems clear that the considerable momentum which HPF possesses, coupled with its attractiveness to customers of parallel computing, will make it an increasing force to be reckoned with in the numerical and scientific computing market. Early versions of HPF are, however, likely to have only limited application areas, since they will have very restricted MIMD support, and neither unstructured data objects nor sophisticated load-balancing support, although such developments are likely in the future. Fortran-90 has been chosen as the base for HPF development, and Fortran-77 will not be supported. This is despite the fact that the US DoD requires Fortran-77 on its contracts, and recommends against the use of Fortran-90. There are also worries that no consideration has been made of providing HPF semantics in C. ================================================================================