home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!sun4nl!cwi.nl!dik
- From: dik@cwi.nl (Dik T. Winter)
- Newsgroups: comp.sys.super
- Subject: Re: World's Most Powerful Computing Sites
- Message-ID: <8699@charon.cwi.nl>
- Date: 21 Jan 93 21:33:18 GMT
- References: <1993Jan20.163946.1607@hubcap.clemson.edu> <1993Jan20.211032.11929@hubcap.clemson.edu> <1993Jan20.232809.29241@nas.nasa.gov>
- Sender: news@cwi.nl
- Organization: CWI, Amsterdam
- Lines: 19
-
- In article <1993Jan20.232809.29241@nas.nasa.gov> fineberg@nas.nasa.gov writes:
- > In article <1993Jan20.211032.11929@hubcap.clemson.edu>, richard@meiko.com (Richard Cownie) writes:
- > |> This seems to imply a figure of over 4MFLOPS per T800. Last time
- > |> I programmed one, it was a real struggle to achieve over 1MFLOPS
- > |> even for inner loops of vector routines.
- ...
- > Its certainly no less realistic than those for the i860.
- >
- I disagree a bit. It is possible on the i860 to get sensible routines
- running at 50 MFLOPS single precision (although also here it is a real
- struggle). But using published peak can be very misleading. I know
- of one machine where the published peak can only be obtained during the
- calculation cycles of the vector instruction. You must ignore instruction
- start-up time. If you included that the peak would be lowered by 25% (yeah,
- right, one quarter). But than we get at the standard question: what should
- you use?
- --
- dik t. winter, cwi, kruislaan 413, 1098 sj amsterdam, nederland
- home: bovenover 215, 1025 jn amsterdam, nederland; e-mail: dik@cwi.nl
-