| 3:09:25 AM PDT - Sat, Apr 27th 2013 |
|
Dear NWChem developers and users,
I am a novice of nwchem. I carried out a DFT task for a system of ~50 atoms using b3lyp/6-311++G**. My calculation parallel run using 24 cores. NWChem program produced some very large scratch files. Each aoints file was larger than 5GB. The CPU usage was very low. The running time of nwchem was much longer compared with Gaussian.
How can I solve this problem to make nwchem faster?
Thanks
|