Previous Thread
Next Thread
Print Thread
Joined: Aug 2013
Posts: 41
Y
Forum Member
OP Offline
Forum Member
Y
Joined: Aug 2013
Posts: 41
I have install the c37a2 charmm with parallelization,and it shows that success,the installation commands is:
./install.com gun xxlarge M

However,when I submit a job, I found the rate is so slow! Even slower than the non-parallelization charmm with just only one CPU!
Besides,I submit the job to the cluster with qsub.and I just use just one node and each node has 8 CPUs.

Joined: Sep 2003
Posts: 4,883
Likes: 12
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 4,883
Likes: 12
c37a2 is not a released version.

Please provide more detail about your benchmark: number of atoms, the input file used to run it, the timing data at the end of the output for both the serial and parallel runs (and for the parallel run it is useful to have the output for all nodes)


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Joined: Sep 2003
Posts: 8,659
Likes: 26
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,659
Likes: 26
The results would seem to argue against a successful compilation, but as noted, the lack of detail makes it difficult to judge or offer suggestions. The only case I can think of where this might happen for correctly compiled code is a test system with a fairly small number of atoms.


Rick Venable
computational chemist


Moderated by  lennart, rmv 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.5
(Release build 20201027)
Responsive Width:

PHP: 7.3.31-1~deb10u5 Page Time: 0.013s Queries: 20 (0.011s) Memory: 0.7372 MB (Peak: 0.7809 MB) Data Comp: Off Server Time: 2023-11-28 18:41:28 UTC
Valid HTML 5 and Valid CSS