From: Sean A. Peebles (sapeebles@eiu.edu)
Date: Wed Jul 09 2003 - 10:05:05 EDT
Thanks to all who replied regarding my memory problem, and to John Ballard
(for his malloc memory test routine). My original problem was that I was
unable to get the Gaussian 98 ab initio program to use more than about
100MB of the physical memory, the program exiting if %mem much greater than
100MB was specified.
The problem was nothing more than the data and stack sizes (checked via
ulimit -a) being set too small - in particular the per_proc_data_size was
set only to the default 128MB. I changed these via dxkerneltuner and I'm
able to get Gaussian to use significantly larger portions of the memory
than I was able to previously.
Thanks again to all who helped out.
Sean.
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:49:26 EDT