home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!olivea!hal.com!dennis.hal.COM!not-for-mail
- From: allison@hal.COM (Dennis Allison)
- Newsgroups: comp.benchmarks
- Subject: Re: Apple mark test
- Message-ID: <159mjqINN8fg@dennis.hal.COM>
- Date: 30 Jul 92 21:23:06 GMT
- References: <1992Jul30.200041.2990@noose.ecn.purdue.edu>
- Organization: HaL Computer Systems, Inc.
- Lines: 16
- NNTP-Posting-Host: dennis.hal.com
-
- In article <1992Jul30.200041.2990@noose.ecn.purdue.edu> shankar@ec.ecn.purdue.edu
- (Shankar Ramakrishnan) writes:
- >
- >The following is posted by me on behalf of Frank Klemm:
- >
- >[ stuff deleted ...] the apple mark test? ... a test [of] how many fractal
- >iteration[s] a computer can make per second. 400 kApple/sec means ... 400000
- >iterations per second. ... [stuff deleted]
- >
- > Now the program (runs only on 386+): [uucoded program deleted]
- >
- Posting a 386 executable isn't much help. It gives a number for
- one architecture/operating system compbination. Cross architecture
- comparisons are clearly precluded. Without the source code no one
- can determine where the bottlenecks lie in a particular machine nor
- can one try different optimization strategies.
-