Greetings everyone,
So I am doing the standard pre-docking protein structure preparation, adding hydrogens giving them plausible bond angles etc.
So I add all the Hydrogens and then create a fixed atom constraint for everything that isn't a hydrogen.
I then type the molecule with charm.
I then try and run the minimisation protocol with processors set to 2 (I'd say 4 but I have another job running and don't want to interrupt it). it clearly runs several steps of preparation (all on a single core) but when it gets to a certain point cpu usage drops to 0 and the job appears to hang. I know it is a problem with trying to use 2 cores as when i run it on one core with all the same settings the job works flawlessly. I am running this on the same machine as the pieline pilot server and it is also our license server, so I can't see this being a passing problem. It also has the same problems on any of our other machines.
Interestingly we can use multiple cores for some protocols, conformer generation being one of them.
Actually all our charm based attempts to use multiple cores fail. (cdocker, MD etc) (we have two accelrys charm licenses specifically so we can do this).
As far as I understand it the second license should only be required for a distinct second CPU not for multiple cores, can anyone confirm/deny?
Thanks in advance.
Tom