home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!uwm.edu!cs.utexas.edu!torn!nott!cunews!ijeff
- From: ijeff@hank.carleton.ca (Ian Jefferson)
- Newsgroups: comp.sys.next.advocacy
- Subject: Re: An idea for improving NeXT performance
- Keywords: performance cpu processors
- Message-ID: <ijeff.727898594@cunews>
- Date: 24 Jan 93 18:03:14 GMT
- Article-I.D.: cunews.ijeff.727898594
- References: <1993Jan22.190646.15919@Princeton.EDU>
- Sender: news@cunews.carleton.ca (News Administrator)
- Organization: Carleton University
- Lines: 22
-
- Yes Yes! but make it more general Purpose than that. I think that thread
- level or even process level multiprocessing is usefull on a desktop
- machine.
-
- I have limited understanding of multiprocessing but I beleive that this
- kind of strategy was employed by Apollo on the DN10000 some years ago.
- In that case the compiler could split tasks off on a procedure by
- procedure basis. The OS scheduler would then assign these light weight
- tasks on a who-is-least-loaded sort of way.
-
- There must be some folks out there with further insight on
- multiprocessing on a NeXT. Isn't mach supposed to do this sort of
- thread level MP? Comments?
-
-
-
- --
- ---------------------------------------------------------------------------
- Ian Jefferson ijeff@ccs.carleton.ca No NeXT mail please!
- ijeff@computeractive.on.ca NeXT mail please!
-