[gmx-users] Simulation runs on iMac but explodes on cluster

[Luke Goodsell]

Hi Emanuel,


What kind of compounds are you simulating?****
>
> ~1800 atom protein


> What size of box are you using?
>
Not sure of specifics, since the tpr file was provided by a collaborator,
but it appears to be ~ 7nm in each dimension


> ****
>
> Do  you run on multiple thread when you run it on your iMac?
>
No. Compiled with --disable-threads


> ****
>
> How many CPU’s are you using on the cluster?
>
1. Compiled with --disable-threads

I'm not so much interested in diagnosing what caused the LINCS failure as
I've spent a lot of time trying that already. I'm more concerned about what
is the difference between the two environments that leads to such a failure.
That one produces a stable simulation and one so stubbornly refuses to work
implies a pretty significant difference.

What can I change to try to identify the cause of this discrepancy?

Thanks,
Luke
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20110713/a3aa4a4c/attachment.html>