Dear Molpro Users: I am running Molpro on part of a cluster (one machine) as part of a molecular dynamics simulation for fromaldehyde. Currently, Molpro is used to calculate gradients of both energy and dipole moments once per step for 5000+ steps.
The program works, but whenever Molpro is called it does not release memory, causing a progressively more serious memory shortage until the machine is rebooted. I would be most grateful if anyone can advise on how to prevent this. The cluster has two Opteron 246 (2 GHz) processors (4 GB memory), but the problem also exists on a local machine with two Pentium 4 (3.00 GHz) processors (1 GB memory). Sincerely, Jesse Kadosh
