On Sunday 24 February 2002 12:22 pm, dave mallery wrote: > the other annoyance... i tried a full run of [EMAIL PROTECTED] > and it took !!16!! hours... that's only a little faster > than a 233mmx. my piii 500's do it in eleven.
FYI, Our $8000 666 MHz UP2000 (21264 Alpha) machine runs a single packet from [EMAIL PROTECTED] in 6 hrs 5 minutes on average. Compare that to our $2000 1.2 GHz AMD Athlon, which runs it in 6 hrs 20 minutes on average. For a little more direct comparison, the same molecular simulation on the UP2000 using the Compaq Fortran for Linux compiler took 4 minutes vs 2 minutes on the Athlon using the NAG compiler. Full optimization was on for both. I was pretty suprised and shocked at the difference between compilers and platforms. Marty Sanborn -- | Martin Sanborn - Dept. of Chemical Engineering - Northwestern University | | [EMAIL PROTECTED] - (847)467-1653 - http://zeolites.cqe.nwu.edu/marty |
