[gmx-users] Relaxation / precision

[Andrew Horsfield]

Hi,

My conjugate gradient relaxation runs stops with:

  Step 61, E-Pot = -2.0995914062e+05, F-max =  1.06133e+04
  Step 62, E-Pot = -2.0994365625e+05, F-max =  1.57691e+05
  Negative w:  -3.571208700467e+21
  z=  -1.243719761920e+11
  gpa=   2.476396229114e+11, gpb=   7.688429179760e+10
  a=   0.000000000000e+00, b=   5.904069366017e-10
  EpotA=  -2.099436562500e+05, EpotB=  -2.098553125000e+05
  Negative number for sqrt encountered (-3571208700466867208192.000000)
  Terminating minimization

Is this a precision problem, or is it something else?

(I cannot get a double precision code to work with PME, which is why I
have not tested this out.)

Cheers,

Andrew