Now I'm very sad: I cannot ask a 7-day job on an IBM-SP6 just to
discover that they may be not enough. Don't you think there's a
bottleneck somewhere in the code? What's the reason for such heaviness?
Rob
On 21/03/2011 18:59, Christopher Rowan wrote:
I performed optical calculations on system
I performed optical calculations on systems with up to 1300 atoms (PVDF
adsorbing onto Ag (30 atoms) on 64 cores. Including all bands and sampling
a 1x1x1 optical mesh the calculations took 7 days.
Chris
On 21 March 2011 03:59, Roberto Guerra wrote:
> Hello,
>
> I'm working on a quite large sys
Hello,
I'm working on a quite large system (Si, O, ~1000 atoms in total) and it
relaxed quite well on 64 opteron cpus. At the end I wish to evaluate the
optical absorption (epsimg) but it takes forever for this computation:
could not complete in 24 hours. As far as I know this calculation shou