Fwd: REMD on IBM iDataPlex DX360M3

From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Tue Jul 08 2014 - 00:41:54 CDT

At CINECA it has been determined that my REMD, with a peptide of 19 aa
under GB conditions, provided too less for the GPUs. In fact, under
explicit water conditions, I could run REMD with also the GPUs. If
anything, I expected more from the GPUs under these conditions: like with
only CPUs, very many replicas are needed in order to have adequate exchange--001a11c2034893ad0b04fda80c37--

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:20:58 CST