home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.parallel
- Path: sparky!uunet!destroyer!gatech!hubcap!fpst
- From: cpr4k@holmes.acc.Virginia.EDU (Christian P. Roberts)
- Subject: Re: CM-2 Performance?
- Message-ID: <1992Nov19.151504.9426@murdoch.acc.Virginia.EDU>
- Sender: usenet@murdoch.acc.Virginia.EDU
- Organization: Academic Computing Center - U.Va.
- References: <1992Nov18.130500.26154@hubcap.clemson.edu>
- Date: Thu, 19 Nov 1992 15:15:04 GMT
- Approved: parallel@hubcap.clemson.edu
- Lines: 28
-
- In article <1992Nov18.130500.26154@hubcap.clemson.edu>, shiraz!ian@ohm.york.ac.uk (I Stephenson) writes:
- |> I've been doing some development work on a CM-2. I've got my software to work
- |> fine, but am unsure as to how efficient my impementation is, compared to other
- |> platforms I've run it on.
- |>
- |> The problem is that all the performance figures I've seen for the CM are in
- |> GFlops. This is all very well but my code is stricly integer. Given that the
- |> CM-2 has dedicated FP units, this tells me very little about the performance I
- |> should be expecting.
- |>
- |> Has anyone got a MIPS rating for the machine, produced using integer code? I
- |> appreciate the limitations of such a metric, but it may give me a rough idea of
- |> whether I'm getting things about right, or missing by a mile! Any other
- |> comments as to the integer performance would be appreciated.
- |>
-
- I tried to email, but it bounced:
-
- Please post a summary of information. I am not using the CM-2 for integer
- work, but I work with clients who do. I would appreciate information about
- this analysis of performance that I could pass on to them.
-
- --
-
- Chris Roberts ITC/Academic Computing Center
- cpr4k@virginia.edu University of Virginia
- (804) 982-4693 Charlottesville, Virginia 22903
-
-