Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!pt.cs.cmu.edu!andrew.cmu.edu!<UNAUTHENTICATED>+
From: mbjr@maxwell.ece.cmu.edu (Mauricio Breternitz)
Newsgroups: comp.arch
Subject: VLIW Architecture - References, other projects
Message-ID: <8909291517.AA00260@maxwell.ece.cmu.edu>
Date: 29 Sep 89 15:17:15 GMT
Reply-To: mbjr@maxwell.ece.cmu.edu (Mauricio Breternitz)
Organization: Carnegie Mellon, Pittsburgh, PA
Lines: 28

Subject: VLIW architecture

You should also look at the IBM VLIW. Some of the main diferences with the
Multiflow are:

	- the architecture is geared towards a different compilation technique
(Percolation Scheduling) which is very effective for branch-intensive code;

	- the conditional-execution model of operation. This also helps
in branch-intensive code.

Some relevant papers are:

"Some Design Ideas for a VLIW Architecture for Sequential Natured
Software"by K.Ebcioglu
Proc. IFIP WG 10.3 Working Conference on Parallel Processing,
M. Cosnard et. al (editors) North Holland, 1988

"A global resource-constrained parallelization technique"
K. Ebcioglu and A.Nicolau, ICS'89, Crete, Greece

"A Development Environment for Horizontal Microcode"
IEEE Transactions on Software Engineering, May 1988

					Mauricio Breternitz
					CMU-Carnegie-Mellon University
					ECE-Electrical and Computer Eng Dept
					mbjr@maxwell.ece.cmu.edu