Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!emory!gatech!hubcap!fpst
From: tve@sprite.berkeley.edu (Thorsten von Eicken)
Newsgroups: comp.parallel
Subject: Re: New i860 parallel machine
Message-ID: <1991Apr18.173441.23902@agate.berkeley.edu>
Date: 18 Apr 91 17:34:41 GMT
References: <1991Apr18.120233.339@hubcap.clemson.edu>
Sender: root@agate.berkeley.edu (Charlie Root)
Organization: University of California, Berkeley
Lines: 9
Approved: parallel@hubcap.clemson.edu

... taking you up on your request for flames...

When I see as new DMMP proposed, the first parameter I look at is how
many cycles it takes to send&receive a message. If it's greater than
10, the machine is not interesting in my opinion. Now, if you try to
glue a few off-the-shelf processors together, you're not likely to
produce anything interesting according to my criterion. But then, maybe
I should only look at peak MFLOPS?
	Thorsten von Eicken - tve@sprite.berkeley.edu

-- 
=========================== MODERATOR ==============================
Steve Stevenson                            {steve,fpst}@hubcap.clemson.edu
Department of Computer Science,            comp.parallel
Clemson University, Clemson, SC 29634-1906 (803)656-5880.mabell