Hi,
I've not done this with NR routines, but I've done a lot of
stuff with F90/5. In my experience, there can be slowdowns
in certain circumstances, if you're not careful. For instance,
there are situations in which the run-time system must create
temporary arrays.
However, the degree of optimization you can expect depends
greatly on the compiler. NAGware is not viewed as a terribly
well-optimizing compiler; it's strength is usually considered
to be early adoption of new features and careful standard
conformance. This is hearsay, and could be wrong; I've
not used it.
You say you're using a Pentium, but you don't say what OS.
(Windoze? Linux? Some other UNIX?) See www.polyhedron.com
for extensive comparative benchmarks of Fortran compilers
for Intel.
I've had very good experiences with Absoft; Lahey and PGF
(Portland Group) are also said to be excellent. CVF
(Compaq Visual Fortran) is also excellent, but it's Windoze only.
-P.
On Wed, 6 Dec 2000, Roland Schilling wrote:
> I was just trying out FFT routines from Numerical Recipes.
> To my great surprise I made the observation that the old
> F77 routine 'realft' (for an array of 1024 real data points)
> is twice as fast as the newer F90 version of it. I compiled
> both with the same NAGWare f95 compiler and run it on a
> 233 MHz Pentium-I PC. I know that the F90 versions are
> written with regard to effective use on parallel processors,
> but didn't expect such a dramatic loss of performance on a
> single processor machine.
>
> Can anybody comment on this or has made similar experience?
>
> Thanks in advance for any reply,
> Roland
>
> PS: This post has also been sent to comp.lang.fortran .
>
>
> +----------------------------------------------------------------------+
> | Roland Schilling Home-Office: +49(89)32929-670 |
> | Max-Planck-Institut fuer Quantenoptik Phone: +49(89)32905-265 |
> | Hans-Kopfermann-Str. 1 Fax: +49(89)32905-200 |
> | D-85748 Garching E-mail: [log in to unmask] |
> | Germany http://www.geo600.uni-hannover.de |
> +----------------------------------------------------------------------+
>
|