On Thu, 20 Sep 2001 15:00:44 -0500
"Kurt Wall" <[EMAIL PROTECTED]> wrote:
| > If you have some relevant, non-proprietary benchmarks for the
| > performance of
| > seismic code in Fortran on Linux vs. C on Linux using an open-source
| > compiler, let's see them. I was unable to find an open source Fortran
| > compiler for Linux - which one are you using?
|
| Um, g77 comes to mind, although I suppose that GNU prefer to call
| g77 a "free" compiler rather than an "open source" compiler.
Although I have not used it, a friend that uses Sun workstations to analyze
MRI data in protein structure research ported his code to g77. The calculation
results he got were different from those on the Sun. Sun's FORTRAN compiler
comes with (or at least has passed tests with) a proper validation suite.
No such suite seems to exist for g77. So, he was forced to continue with the
Sun compiler. There are some settings for g77 that may have influenced this.
However, determining which g77 settings allow conformity with Sun's FORTRAN
compiler was not an obvious task. This was 2-3 years ago and g77 has changed
since then. Maybe this is no longer an issue.
--
=====================================================================
Roger Oberholtzer E-mail: [EMAIL PROTECTED]
OPQ Systems AB WWW: http://www.opq.se
Erik Dahlbergsgatan 41-43 Phone: Int + 46 8 314223
115 32 Stockholm Mobile: Int + 46 733 621657
Sweden Fax: Int + 46 8 302602
_______________________________________________
http://linux.nf -- [EMAIL PROTECTED]
Archives, Subscribe, Unsubscribe, Digest, Etc
->http://linux.nf/mailman/listinfo/linux-users
