home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.benchmarks
- Path: sparky!uunet!gumby!destroyer!news.iastate.edu!help.cc.iastate.edu!willmore
- From: willmore@iastate.edu (David Willmore)
- Subject: Re: DEC ALPHA Performance Claims
- Message-ID: <willmore.722118562@help.cc.iastate.edu>
- Sender: news@news.iastate.edu (USENET News System)
- Organization: Iowa State University, Ames IA
- References: <BxH7s7.5Cv@inews.Intel.COM> <4248@bcstec.ca.boeing.com> <BxuCsv.FwE@pix.com>
- Date: Wed, 18 Nov 1992 20:29:22 GMT
- Lines: 25
-
- stripes@pix.com (Josh Osborne) writes:
- >silverm@bcstec.ca.boeing.com (Jeff Silverman) writes:
- >[...]
- >>1) Newer implementations of the alpha may need or take advantage of new ways
- >>of ordering instructions. This in turn means that different alpha
- >>implementations may need or take advantage of different executables or object
- >>modules, each compiled for its own processor. Or maybe the processors will
- >>take a performance hit in the name of portability. Who can say?
-
- >That's not true of only the Alpha ('tho thay may make no atttempt to lessen
- >the blow). One way to solve this problem is to use something like OSF's
- >ANDF. I don't know if it is a good way (it has lots of other advantages and
- >disadvantages), but it should work.
- >[...]
-
- I don't expect that will be much of a problem with the Alpha. From the way the
- archetecture is designed, it looks like the most optimization and ordering takes
- place on-chip. The compiler orders things in a fairly general way which the CPU
- schedules. I wouldn't say that Alpha eliminates all code ordering problems, but
- it should be less effected than other archetectures. At least I hope so.
-
- ---------------------------------------------------------------------------
- willmore@iastate.edu | "Death before dishonor" | "Better dead than greek" |
- David Willmore | "Ever noticed how much they look like orchids? Lovely!" |
- ---------------------------------------------------------------------------
-