Subject: [gmx-users] NVE problems

[Michael Shirts]

Hi, Istvan-

Could you forward the .mdp you used to the list?  1 ns should be sufficient
for an unconstrained system to avoid bad numerical instability (though there
will be some heating) IF the right .mdp options (which are NOT the default
options are used).  If you can e-mail the .mdp, I'll see if I can pick out
what might be causing problems.

Cheers,
Michael Shirts
Stanford University