[SciPy-dev] More info on 100x100x100 FDTD's with Weave

rob rob at pythonemproject.com
Wed Mar 27 09:12:23 CST 2002


The more info is that there is no more info that makes sense.  Verbose=2
gave me nothing.  When I switch from 80 cubed to 100 cubed region, as
soon as the program hits the first Weave statement, my box starts paging
with no CPU activity.  It finally maxes out at 300 Meg swap and kills
the app.  No verbose debugging info to be found.  Plain Numpy works fine
with this sim at 300Meg+ memory.  It does require a lot of swap on my
HD, but there is plenty left to run Netscape, X11, etc.

I have posted the program on my web site.  Hit the "Whats New" button. 
Its called UFOblitz.py since I am doing a quasi-static analysis of those
HV flying craft that were recently on Slashdot.  If someone else can
test this program and prove my results I will be happy.  Otherwise this
might be a FreeBSD problem.  Just run the program normally, then run
again after changing the "cells/meter" variable to 100.  (Its in a 1M
cubed sim space)

Thanks,  Rob.



-- 
-----------------------------
The Numeric Python EM Project

www.pythonemproject.com



More information about the Scipy-dev mailing list