home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!spool.mu.edu!sdd.hp.com!uakari.primate.wisc.edu!ames!agate!boulder!news!grunwald
- From: grunwald@foobar.cs.colorado.edu (Dirk Grunwald)
- Subject: Re: Multi threaded CPUs
- In-Reply-To: schow@bqneh23.bnr.ca's message of Thu, 3 Sep 1992 14:33:41 GMT
- Message-ID: <GRUNWALD.92Sep3201739@foobar.cs.colorado.edu>
- Lines: 17
- Sender: news@colorado.edu (The Daily Planet)
- Nntp-Posting-Host: foobar.cs.colorado.edu
- Reply-To: grunwald@foobar.cs.colorado.edu
- Organization: University of Colorado at Boulder
- References: <Sep02.141909.45739@yuma.ACNS.ColoState.EDU>
- <1992Sep3.143341.11285@bcars64a.bnr.ca>
- Date: Fri, 4 Sep 1992 02:17:45 GMT
- Lines: 17
-
- >>>>> On Thu, 3 Sep 1992 14:33:41 GMT, schow@bqneh23.bnr.ca (Stanley T.H. Chow) said:
- STC> Interesting. I had no idea that Motorola had any multi-threaded
- STC> CPU. I understood that another company (who dominates the PC market,
- STC> but will remain nameless :-) actually have at least one CPU for sale
- STC> with multithreading used internally but not visible to the
- STC> customer.
- --
-
- This would be difficult, I would think. Harry Jordan at Univ. Colorado
- told me he's had students build multithreaded micro cores as class
- projects. It's not difficult, and effectively hides memory latency.
- The question is, when will this become commercially viable?
-
- I think the answer rhymes with ``when will offchip memory delays be
- more than 25 cycles''.
-
- Dirk Grunwald Asst. Prof, Univ. of Colorado at Boulder
-