Sorry, something gone wrong and message was refused. Next try...
#####
Hello Chris, thanks a lot for the suggestions.
> i noticed you are not using any type of optimization in your
> compile.txt...
> Try
> g77 -fno-automatic -Wall -O6 nec2dxs.f -o nec2dxsXXX.exe
>
> i tested with "t-ant.nec" distributed with 4nec2v5 (actually the .inp
> file generated by 4nec2 5.0)
>
> using what was in your compile.txt....
> RUN TIME = 24.084
> using the above...
> RUN TIME = 14.440
> using jsut the version with 4nec2v5...
> RUN TIME = 13.340
>
I do not quite understand...
The first time I presume you used the Nec2dXS.exe compiled as described
in my compile.txt
The second time I presume you used the same source but with one or more
optimization flags.
However I can not place the third run. If the Nec2dXS*.exe belonging to
the 4nec2 package was used it should give the same results as the first
run because they both should be the same.
> also noticed you have dataflow-type compile warnings generated once
> optimization is turned on. these warnings can only be generated once
> the optimizer is on... these might be doing bad things to your
> numerics...
I (and some other users) did a fair number of tests to compare the
original nec2d*.exe results with the nec2dXS*.exe results, but found
no particular differences.
I'll see what I can do with this new warning info and check out
the new version.
Arie.
>
> --
> (Please ignore the advertisements below)
>
> +++ GMX - Mail, Messaging & more http://www.gmx.net +++
> NEU: Mit GMX ins Internet. Rund um die Uhr für 1 ct/ Min. surfen!
>
>
-- (Please ignore the advertisements below) +++ GMX - Mail, Messaging & more http://www.gmx.net +++ NEU: Mit GMX ins Internet. Rund um die Uhr für 1 ct/ Min. surfen! -- The NEC-List mailing list <nec-list_at_gweep.ca> http://www.gweep.ca/mailman/listinfo.cgi/nec-listReceived on Thu Jan 16 2003 - 22:21:10 EST
This archive was generated by hypermail 2.2.0 : Sat Oct 02 2010 - 00:10:44 EDT