[gmx-users] constraint is serial and parallel runs

[Dr. Vitaly Chaban]

Dear All -

Can anyone please explain in simple words why the [LINCS] constraints,
which are stable in the serial run, tend to crash the parallel
simulation. The problem itself can be fixed by using smaller
time-step, but the challenge is to understand the pathway through
which it happens. One basically assumes that interactions are the
same, despite the number of cores involved.

Thank you very much,

Dr. Vitaly V. Chaban
MEMPHYS - Center for Biomembrane Physics
Department of Physics, Chemistry and Pharmacy
University of Southern Denmark
Campusvej 55, 5230 Odense M, Denmark