<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hi,<br>
<br>
On 11/24/10 2:59 AM, lin hen wrote:
<blockquote cite="mid:BLU156-w502393662E51E7636780F4AF3F0@phx.gbl"
type="cite">
<meta http-equiv="Content-Type" content="text/html;
charset=ISO-8859-1">
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style>
hi,<br>
<br>
I am testing the dhfr benchmarks with gromacs 4.5.2<br>
<br>
1. for the benchmarks: dhfr-solv-RF-1nm.bench and
dhfr-solv-RF-2nm.bench, set the steps = 10000, <br>
The GPU version running results shows as following:<br>
./mdrun-gpu<br>
<br>
Pre-simulation ~15s memtest in progress...done, no errors
detected<br>
starting mdrun 'Protein in water'<br>
10000 steps, 20.0 ps.<br>
Writing final coordinates.<br>
Back Off! I just backed up confout.gro to ./#confout.gro.2#<br>
Post-simulation ~15s memtest in progress...done, no errors
detected<br>
OpenMM run - timing based on wallclock.<br>
<strong> NODE (s) Real (s) (%)<br>
Time: 90.796 90.796 100.0<br>
1:30<br>
(Mnbf/s) (MFlops) (ns/day) (hour/ns)<br>
Performance: 0.000 0.060 19.034 1.261</strong><br>
gcq#100: "Jesus Can't Save You, Though It's Nice to Think He
Tried" (Black Crowes)<br>
<br>
The CPU version running results shows as following:<br>
./mdrun<br>
<br>
Writing final coordinates.<br>
Back Off! I just backed up confout.gro to ./#confout.gro.3#<br>
Average load imbalance: 4.7 %<br>
Part of the total run time spent waiting due to load imbalance:
2.7 %<br>
Steps where the load balancing was limited by -rdd, -rcon and/or
-dds: X 0 % Y 0 % Z 0 %<br>
<br>
Parallel run - timing based on wallclock.<br>
<strong> NODE (s) Real (s) (%)<br>
Time: 86.176 86.176 100.0<br>
1:26<br>
(Mnbf/s) (GFlops) (ns/day) (hour/ns)<br>
Performance: 741.634 35.849 20.054 1.197</strong><br>
gcq#0: Thanx for Using GROMACS - Have a Nice Day<br>
<br>
<br>
The gpu version looks <strong>slower </strong>than cpu version,
does it make sense? Is there anyway I could optimize?<br>
<br>
</blockquote>
<br>
In the default build, the CPU version will use all available cores
on the machine and it's faster. <br>
<br>
<blockquote cite="mid:BLU156-w502393662E51E7636780F4AF3F0@phx.gbl"
type="cite">
<br>
2. For the CPU benchmark dhfr-impl-1nm.bench, when I run : <br>
./mdrun<br>
Back Off! I just backed up md.log to ./#md.log.9#<br>
Reading file topol.tpr, VERSION 4.5.1-dev-20100917-b1d66 (single
precision)<br>
Starting 16 threads<br>
-------------------------------------------------------<br>
Program mdrun, VERSION 4.5.2<br>
Source code file: domdec.c, line: 6428<br>
Fatal error:<br>
There is no domain decomposition for 16 nodes that is compatible
with the given box and a minimum cell size of 1.02425 nm<br>
Change the number of nodes or mdrun option -rcon or -dds or your
LINCS settings<br>
Look in the log file for details on the domain decomposition<br>
For more information and tips for troubleshooting, please check
the GROMACS<br>
website at <a moz-do-not-send="true"
href="http://www.gromacs.org/Documentation/Errors">http://www.gromacs.org/Documentation/Errors</a><br>
-------------------------------------------------------<br>
Thanx for Using GROMACS - Have a Nice Day<br>
<br>
<br>
Where could I modify this nodes#? <br>
</blockquote>
<br>
Give the option -nt to mdrun to specify the number of threads you
want to launch, e.g.<br>
<br>
$ mdrun -nt 8<br>
<br>
<blockquote cite="mid:BLU156-w502393662E51E7636780F4AF3F0@phx.gbl"
type="cite">
<br>
3. for CPU benchmark dhfr-impl-inf.bench<br>
I modified cpu-imp-RF-inf.mdp: nsteps 10000<br>
grompp -f cpu-imp-RF-inf.mdp -o topol.tpr<br>
it will show: <br>
NOTE 1 [file cpu-imp-RF-inf.mdp]:<br>
The Berendsen thermostat does not generate the correct kinetic
energy<br>
distribution. You might want to consider using the V-rescale
thermostat.<br>
Generated 2278 of the 2278 non-bonded parameter combinations<br>
Generating 1-4 interactions: fudge = 0.5<br>
Generated 2278 of the 2278 1-4 parameter combinations<br>
Excluding 3 bonded neighbours molecule type 'Protein'<br>
turning all bonds into constraints...<br>
NOTE 2 [file topol.top, line 23568]:<br>
System has non-zero total charge: -1.100000e+01<br>
<br>
GB parameter(s) missing or negative for atom type 'H0'<br>
-------------------------------------------------------<br>
Program grompp, VERSION 4.5.1<br>
Source code file: grompp.c, line: 870<br>
Fatal error:<br>
Can't do GB electrostatics; the forcefield is missing 1 values for<br>
atomtype radii, or they might be negative<br>
.<br>
For more information and tips for troubleshooting, please check
the GROMACS<br>
website at <a moz-do-not-send="true"
href="http://www.gromacs.org/Documentation/Errors">http://www.gromacs.org/Documentation/Errors</a><br>
-------------------------------------------------------<br>
"Everybody's Good Enough For Some Change" (LIVE)<br>
<br>
and when I run ./mdrun<br>
the steps still -1, infinite time<br>
How could I genete the right topol.tpr?<br>
<br>
</blockquote>
Did you search the mailing lists? <br>
<br>
Copy and paste the error message in
<a class="moz-txt-link-freetext" href="http://www.gromacs.org/Support/Mailing_Lists/Search">http://www.gromacs.org/Support/Mailing_Lists/Search</a><br>
and you will get a top hit<br>
<a class="moz-txt-link-freetext" href="http://lists.gromacs.org/pipermail/gmx-users/2010-September/053764.html">http://lists.gromacs.org/pipermail/gmx-users/2010-September/053764.html</a><br>
<br>
You are using grompp 4.5.1 and mdrun 4.5.2 !? Remove all gromacs
installs that you have, make a clean 4.5.3 install and try again.<br>
<br>
Rossen<br>
<br>
<blockquote cite="mid:BLU156-w502393662E51E7636780F4AF3F0@phx.gbl"
type="cite">
<br>
Thanks a lot.<br>
<br>
<br>
YY<br>
</blockquote>
<br>
</body>
</html>