Seeking vectorizer/parallelizer

Michael Ernst <mernst@theory.lcs.mit.edu>
Fri, 26 Jun 1992 23:23:44 GMT

          From comp.compilers

Related articles
Seeking vectorizer/parallelizer mernst@theory.lcs.mit.edu (Michael Ernst) (1992-06-26)
Parallelizer and vectorizers available mernst@theory.lcs.mit.edu (Michael Ernst) (1992-07-13)
Re: Parallelizer and vectorizers available paco@cs.rice.edu (1992-07-13)
FORGE 90 vectorizer rchrd@netcom.com (1992-07-14)
Re: Parallelizer and vectorizers available paco@cs.rice.edu (Paul Havlak) (1992-07-15)
| List of all articles for this month |
Newsgroups: comp.parallel,comp.compilers
From: Michael Ernst <mernst@theory.lcs.mit.edu>
Organization: MIT Lab for Computer Science
Date: Fri, 26 Jun 1992 23:23:44 GMT
Keywords: parallel, optimize, vector

I am looking for vectorizers or parallelizers for a research project. (I
have a scheme for translating parallel code into sequential and I want to
see whether I can translate their code into efficient sequential
implementations and vice-versa.) I'm most interested in publicly
available tools because I don't have a lot of money to spend on a tool
I'll only be using for a few months. I'm not picky about the
parallelizer's target language: Fortran, C, or any similar imperative
language is fine with me.


I know about Pat (Georgia Tech interactive source-to-source Parallelizing
Assistant Tool; peabody.gatech.edu:pub/pat) and the Omega test (University
of Maryland array data dependence analyzer; ftp.cs.umd.edu:pub/omega). I
haven't yet tried either but can give pointers to the code to those
interested. I also know about the commercial products KAP (Kuck &
Associates) and VAST (Pacific-Sierra Research), but don't have addresses
for either company and assume that they wouldn't fit my other needs. I
seem to recall a posting about a service where you emailed your code
somewhere and it came back with data dependences marked, but I can no
longer find a reference to it.


Other pointers would be greatly appreciated; I will be happy to summarize
to interested parties.
-Michael Ernst
mernst@theory.lcs.mit.edu
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.