home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.lang.fortran
- Path: sparky!uunet!wupost!gumby!destroyer!ubc-cs!unixg.ubc.ca!kakwa.ucs.ualberta.ca!acs.ucalgary.ca!bauwens
- From: bauwens@acs.ucalgary.ca (Luc Bauwens)
- Subject: Re: MS Fortran 5.1 upgrade. Is it worth it?
- Sender: news@acs.ucalgary.ca (USENET News System)
- Message-ID: <92Jul26.023952.18001@acs.ucalgary.ca>
- Date: Sun, 26 Jul 92 02:39:52 GMT
- References: <1992Jul25.220659.2257@news.Hawaii.Edu>
- Nntp-Posting-Host: acs5.acs.ucalgary.ca
- Organization: The University of Calgary, Alberta
- Lines: 27
-
- In article <1992Jul25.220659.2257@news.Hawaii.Edu> tholen@galileo.ifa.hawaii.edu (Dave Tholen) writes:
- >Luc Bauwens writes:
- >
- >> The 32 bit apps run about 35% faster than the 16 bits MS apps.
- >
- >I have found that the speedup depends on the relative amounts of I/O and
- >number crunching. WATCOM's I/O performance would appear to be slightly
- >worse than Microsoft's, but in the number-crunching department, the
- >32-bit code wins handily, and by more than a factor of two.
-
-
- Not that I doubt your results, but my observation of a 35% speed increase
- was with very little i/o, primarily double-precision floating point
- stuff.
-
- Potential explanations for the differences? Somebody suggested me
- that it me be related to array size: my code does not contain any
- array larger than 64 kB.
-
- Also, there is a number of transcendental functions in my code, which
- may not benefit from the improvement?
-
- (And perhaps I should do some kind of sensitivity analysis, just
- change my array sizes and look at what happens?)
-
- Luc B
-
-