Dear MOLPRO users,
I am currently trying to get numerical frequencies at the (canonical) CCSD(T)-F12 level for somewhat larger molecules and I am reaching the limits of our computational resources on single nodes. I would therefore like to run the calculations across several nodes, which I guess should be possible using the mppx mode (every displacement is handled on a single CPU). However, I run into trouble because the program tries to perform the initial calculation of the procedure parallelized across all nodes, which fails (already rhf gives the error ``Program RHF-SCF not suitable for multi-node calculations´´ using Version 2024.1). Is there a way to suppress the parallelization in the initial execution of the procedure or to treat it as a displacement so that it is performed using one core only? Or am I just missing a simpler solution?
Thanks in advance for the help!
Best,
Marc