[gmx-users] Number of nodes
Mark.Abraham at anu.edu.au
Mon Oct 31 13:42:31 CET 2011
On 31/10/2011 11:29 PM, Kavyashree M wrote:
> System Admin said that the Job fails on 32 and 128 because memory
> is insufficient for each task, so upon increasing the nodes, data gets
> distributed across more number of nodes and each node gets less
> memory occupancy and he also mentioned that he was able to run
> on 512 nodes but it was giving error for not having sufficient data for
> 512 nodes.
That combination of observations is inconceivable. 60K atom simulations
run just fine on 32 nodes in coprocessor or virtual-node mode, and I'd
bet 600K atoms is also fine.
> I have run the same job on 8 nodes in i7 machine.
A very rough rule of thumb is that at least about a thousand atoms per
processor is enough to make that number of processors worth using.
> Thank you
> With Regards
> On Mon, Oct 31, 2011 at 5:47 PM, Justin A. Lemkul <jalemkul at vt.edu
> <mailto:jalemkul at vt.edu>> wrote:
> Kavyashree M wrote:
> Dear Users,
> I was trying to run a simulation (gromacs4.5.3)
> on a Bluegene/L machine. But I was unable to run.
> System admin say that I need to change the input
> file. I am not sure what needs to be changed in the
> input file which specifies no. of nodes usage.
> Sounds like your system admin should be offering you some more
> help, since they know the specifics of the hardware, queuing
> software, etc.
> I am not familiar with the bluegene machines. Kindly
> suggest the possible solutions.
> Without seeing your input file(s), there's nothing anyone can do.
> Be mindful that this list pertains to Gromacs-specific problems,
> which do not seem to be present here.
> Justin A. Lemkul
> Ph.D. Candidate
> ICTAS Doctoral Scholar
> MILES-IGERT Trainee
> Department of Biochemistry
> Virginia Tech
> Blacksburg, VA
> jalemkul[at]vt.edu <http://vt.edu> | (540) 231-9080
> gmx-users mailing list gmx-users at gromacs.org
> <mailto:gmx-users at gromacs.org>
> Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> Please don't post (un)subscribe requests to the list. Use the www
> interface or send it to gmx-users-request at gromacs.org
> <mailto:gmx-users-request at gromacs.org>.
> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the gromacs.org_gmx-users