jianmu
/
lammps
forked from lijiext/lammps
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@3283 f3b2605a-c512-4ea7-a41b-209d697bcdaa

This commit is contained in:
sjplimp 2009-11-02 17:38:32 +00:00
parent f578c6823a
commit acb41c8e79
2 changed files with 46 additions and 40 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
doc/prd.html
View File

@ -53,27 +53,30 @@ prd 5000 100 10 10 100 1 54982 maxiter 100
</PRE>
<P><B>Description:</B>
</P>
<P>Run Parallel Replica Dynamics (PRD) as described in <A HREF = "#Voter">(Voter)</A>.
PRD is a method for performing accelerated dynamics that is suitable
for infrequent-event systems that obey first-order kinetics. To quote
from the paper: "The dynamical evolution is characterized by
vibrational excursions within a potential basin, punctuated by
occasional transitions between basins." The transition probability is
characterized by p(t) = k*exp(-kt) where k is the rate constant.
<P>Run Parallel Replica Dynamics (PRD) as described in <A HREF = "#Voter">Art Voter's
paper</A>. PRD is a method for performing accelerated dynamics
that is suitable for infrequent-event systems that obey first-order
kinetics. To quote from the paper: "The dynamical evolution is
characterized by vibrational excursions within a potential basin,
punctuated by occasional transitions between basins." The transition
probability is characterized by p(t) = k*exp(-kt) where k is the rate
constant.
</P>
<P>A PRD run is performed by running independent simulations on multiple
replicas of the same system. To run with M replicas, you must launch
LAMMPS on M partitions, where a partition is one or more processors.
This is done by using the "-partition" command-line argument when
LAMMPS is launched. See <A HREF = "Section_start.html#2_6">this section</A> of the
manual for details. A PRD run can be performed on a single partition,
though this offers no effective parallel speed-up in searching for
infrequent events.
replicas of the same system, which gives an effective enhancement in
the timescale spanned by the multiple simulations, waiting for an
event to occur. To run with M replicas, you must launch LAMMPS on M
partitions, where a partition is one or more processors. This is done
by using the "-partition" command-line argument when LAMMPS is
launched. See <A HREF = "Section_start.html#2_6">this section</A> of the manual for
details. A PRD run can be performed on a single partition, though
this offers no effective parallel speed-up in searching for infrequent
events.
</P>
<P>When a PRD run is performed, it is assumed that each replica is
running the same model, though LAMMPS does not check that this is the
case. I.e. the simulation domain, the number of atoms, the
interaction potentials, etc are the same for every replica.
running the same model, though LAMMPS does not check for this.
I.e. the simulation domain, the number of atoms, the interaction
potentials, etc are the same for every replica.
</P>
<P>A PRD run has several stages, which are repeated each time an "event"
occurs in one of the replicas, as defined below. The logic for a PRD
@ -93,9 +96,9 @@ run is as follows:
</PRE>
<P>Before this loop begins, the state of the system on replica 0 is
shared with all replicas, so that all replicas begin from the same
initial state. The first basin is identified by quenching (an energy
minimization, see below) the initial state and storing the resulting
coordinates for reference.
initial state. The first potential energy basin is identified by
quenching (an energy minimization, see below) the initial state and
storing the resulting coordinates for reference.
</P>
<P>In the first stage, dephasing is performed by each replica
independently to eliminate correlations between replicas. This is

43
doc/prd.txt
View File

@ -40,27 +40,30 @@ prd 5000 100 10 10 100 1 54982 maxiter 100 :pre
[Description:]
Run Parallel Replica Dynamics (PRD) as described in "(Voter)"_#Voter.
PRD is a method for performing accelerated dynamics that is suitable
for infrequent-event systems that obey first-order kinetics. To quote
from the paper: "The dynamical evolution is characterized by
vibrational excursions within a potential basin, punctuated by
occasional transitions between basins." The transition probability is
characterized by p(t) = k*exp(-kt) where k is the rate constant.
Run Parallel Replica Dynamics (PRD) as described in "Art Voter's
paper"_#Voter. PRD is a method for performing accelerated dynamics
that is suitable for infrequent-event systems that obey first-order
kinetics. To quote from the paper: "The dynamical evolution is
characterized by vibrational excursions within a potential basin,
punctuated by occasional transitions between basins." The transition
probability is characterized by p(t) = k*exp(-kt) where k is the rate
constant.
A PRD run is performed by running independent simulations on multiple
replicas of the same system. To run with M replicas, you must launch
LAMMPS on M partitions, where a partition is one or more processors.
This is done by using the "-partition" command-line argument when
LAMMPS is launched. See "this section"_Section_start.html#2_6 of the
manual for details. A PRD run can be performed on a single partition,
though this offers no effective parallel speed-up in searching for
infrequent events.
replicas of the same system, which gives an effective enhancement in
the timescale spanned by the multiple simulations, waiting for an
event to occur. To run with M replicas, you must launch LAMMPS on M
partitions, where a partition is one or more processors. This is done
by using the "-partition" command-line argument when LAMMPS is
launched. See "this section"_Section_start.html#2_6 of the manual for
details. A PRD run can be performed on a single partition, though
this offers no effective parallel speed-up in searching for infrequent
events.
When a PRD run is performed, it is assumed that each replica is
running the same model, though LAMMPS does not check that this is the
case. I.e. the simulation domain, the number of atoms, the
interaction potentials, etc are the same for every replica.
running the same model, though LAMMPS does not check for this.
I.e. the simulation domain, the number of atoms, the interaction
potentials, etc are the same for every replica.
A PRD run has several stages, which are repeated each time an "event"
occurs in one of the replicas, as defined below. The logic for a PRD
@ -80,9 +83,9 @@ while (time remains):
Before this loop begins, the state of the system on replica 0 is
shared with all replicas, so that all replicas begin from the same
initial state. The first basin is identified by quenching (an energy
minimization, see below) the initial state and storing the resulting
coordinates for reference.
initial state. The first potential energy basin is identified by
quenching (an energy minimization, see below) the initial state and
storing the resulting coordinates for reference.
In the first stage, dephasing is performed by each replica
independently to eliminate correlations between replicas. This is