A set of biomolecular systems is presented, which can be used to benchmark
the performance of simulation programs and computers. It is applied, using
the GROMOS96 biomolecular simulation software, to a variety of computers. T
he dependence of computing time on a number of model and computational para
meters is investigated. An extended pair list technique to select non-bonde
d interaction pairs and long-range interactions is shown to increase the ef
ficiency by a factor 1.5 to 3 when compared to standard procedures. The ben
chmark results can be used to estimate the computer time required for simul
ation studies, and to evaluate the efficiency of various computers regardin
g molecular simulations. (C) 2000 Elsevier Science B.V. All rights reserved
.