Xref: utzoo comp.arch:4167 comp.sys.misc:1318 comp.misc:2196
Path: utzoo!mnetor!uunet!lll-winken!lll-tis!ames!amdcad!cae780!leadsv!laic!darin
From: darin@laic.UUCP (Darin Johnson)
Newsgroups: comp.arch,comp.sys.misc,comp.misc
Subject: Any experience/comments with Multiflow machines?
Message-ID: <191@laic.UUCP>
Date: 30 Mar 88 17:47:28 GMT
Organization: Lockheed AI Center, Menlo Park
Lines: 18
Keywords: multiflow parallel compiler

Does anyone have any comments or experience with the Multiflow machines?
In particular, anything "bad" that didn't come in their sales promo.

Here are my first assumptions, correct me if I am off target:

  + able to make use of existing code (dusty decks) and parallelize it
      without code modification (ie. transforming DO loops, etc.)
  + inexpensive
  + will run faster than Alliant, Convex computers for programs
      that aren't vectorizable or conducive to their optimizing
      compilers
  - will run slower than above machines if programs are highly
      vectorizable and/or are modified to take advantage of
      the target machine.

-- 
Darin Johnson (...ucbvax!sun!sunncal!leadsv!laic!darin)
              (...lll-lcc.arpa!leadsv!laic!darin)
	All aboard the DOOMED express!