> John,
>
> I've been running NEC4 and HFSS (finite element code) for a few years on
> various computers. I recently upgraded to a Dell Precision 620 Workstation
> with dual 1GHz Pentium III processors, 2GBytes of RAM, and 15K RPM hard
> drives (RAID capability is available). I'm running on NT 4.0 due to
> limitations in my application software, but will be changing to Windows 2000
> soon.
>
> This system was recommended to me by engineers at Northrop Grumman in
> Melbourne and by application engineers at Ansoft, the developers of HFSS
> (both of these locations have purchased the 620 Workstations for their
> engineers).
>
> The total cost was under $10K and I've been very happy with the performance.
> I don't have a good feel for speed improvements you could expect, but if you
> have a sample NEC problem, I'd be happy to run it on my machine for you and
> send back the .nou file with processing time.
Thanks for replying, Eric. I don't know what FORTRAN compiler you are
using for NEC4 but does it offer any option/optimization to use your
dual processor environment? I am also looking at 64-bit processors
since I do a lot of double precision floating point calculations in
NEC. I wish there was an easy way to know prior to purchase the
processing time savings that I might get with a 64-bit vice 32-bit
architecture.
I appreciate your willingness to run a benchmark. Are you set up to
handle a 20,000 segment problem with 4-fold symmetry?
-- __ __ ______ __ John Wood / |\ / /\ / ____ \ / /\ Code 5551 / | ||/ / / / /\__/ /| / / / U.S. Naval Research Lab / /| |/ / / / /_/_/ / / / / / 4555 Overlook Avenue, SW / / | / / / ___ / / / / / Washington, DC 20375-5337 / / /| / / / /\_| |\_/ / /_/_ (202) 767-2608 /_/ / |_/ / /_/ / |_|| /_____/\ (202) 767-3377 (FAX) \_\/ \_\/ \_\/ \_\| \_____\/ e-mail: wood_at_itd.nrl.navy.mil WWW: http://server5550.itd.nrl.navy.mil NEVER EAT MORE THAN YOU CAN LIFT. - Miss PiggyReceived on Fri Feb 09 2001 - 14:28:32 EST
This archive was generated by hypermail 2.2.0 : Sat Oct 02 2010 - 00:10:41 EDT