[gmx-users] T-coupling fixed...

[Senthil Kandasamy]

Hi all, 

This is with reference to the two earlier emails that I had sent.
I recompiled fftw and gromacs using cc and then tried the same
simulations and the temperature coupling seems normal . (I have run a
couple of simulations for ~25 ps and the log files show reasonable
temperature values :310 +- 2 ). So the problem seems to be an earlier
compilation of gromacs with pgcc.(Portland group)

This raises some serious issues. The gromacs version compiled without
failing with pgcc. There was a bunch of warnings, but  I kind of ignored
them assuming they were not that important since it did ultimately
compile.As far as I can tell, the only weird behavior was in the
Temperature coupling,(and possibly pressure coupling, since the box was
flattened a little, more than the typical anisotropic coupling) 
specifically with respect to water. The developers need to look into
this, I think...


Senthil