ICODE-ADC/lammps
4
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

''

Browse Source 
git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@13961 f3b2605a-c512-4ea7-a41b-209d697bcdaa
This commit is contained in:
sjplimp
2015-08-29 00:13:36 +00:00
parent f945c6d79a
commit 4d13f3d33d
1 changed files with 23 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

45
doc/doc2/Section_start.html
View File

@ -1753,9 +1753,10 @@ thermodynamic state and a total run time for the simulation. It then
appends statistics about the CPU time and storage requirements for the
simulation. An example set of statistics is shown here:
</P>
<PRE>Loop time of 2.81192 on 4 procs for 300 steps with 2004 atoms
97.0% CPU use with 4 MPI tasks x no OpenMP threads
Performance: 18.436 ns/day 1.302 hours/ns 106.689 timesteps/s
<P>Loop time of 2.81192 on 4 procs for 300 steps with 2004 atoms
</P>
<PRE>Performance: 18.436 ns/day 1.302 hours/ns 106.689 timesteps/s
97.0% CPU use with 4 MPI tasks x no OpenMP threads
</PRE>
<PRE>MPI task timings breakdown:
Section | min time | avg time | max time |%varavg| %total
@ -1783,16 +1784,16 @@ Neighbor list builds = 26
Dangerous builds = 0
</PRE>
<P>The first section provides a global loop timing summary. The loop time
is the total wall time for the section. The second line provides the
CPU utilzation per MPI task; it should be close to 100% times the number
of OpenMP threads (or 1). Lower numbers correspond to delays due to
file i/o or unsufficient thread utilization. The <I>Performance</I> line is
is the total wall time for the section. The <I>Performance</I> line is
provided for convenience to help predicting the number of loop
continuations required and for comparing performance with other similar
MD codes.
continuations required and for comparing performance with other
similar MD codes. The CPU use line provides the CPU utilzation per
MPI task; it should be close to 100% times the number of OpenMP
threads (or 1). Lower numbers correspond to delays due to file I/O or
insufficient thread utilization.
</P>
<P>The second section gives the breakdown of the CPU run time (in seconds)
into major categories:
<P>The MPI task section gives the breakdown of the CPU run time (in
seconds) into major categories:
</P>
<UL><LI><I>Pair</I> stands for all non-bonded force computation
<LI><I>Bond</I> stands for bonded interactions: bonds, angles, dihedrals, impropers
@ -1811,17 +1812,17 @@ the difference between minimum, maximum and average is small and thus
the variation from the average close to zero. The final column shows
the percentage of the total loop time is spent in this section.
</P>
<P>When using the <A HREF = "timers.html">timers full</A> setting, and additional column
is present that also prints the CPU utilization in percent. In addition,
when using <I>timers full</I> and the <A HREF = "package.html">package omp</A> command are
active, a similar timing summary of time spent in threaded regions to
monitor thread utilization and load balance is provided. A new enrty is
the <I>Reduce</I> section, which lists the time spend in reducing the per-thread
data elements to the storage for non-threaded computation. These thread
timings are taking from the first MPI rank only and and thus, as the
breakdown for MPI tasks can change from MPI rank to MPI rank, this
breakdown can be very different for individual ranks. Here is an example
output for this optional output section:
<P>When using the <A HREF = "timers.html">timers full</A> setting, an additional column
is present that also prints the CPU utilization in percent. In
addition, when using <I>timers full</I> and the <A HREF = "package.html">package omp</A>
command are active, a similar timing summary of time spent in threaded
regions to monitor thread utilization and load balance is provided. A
new entry is the <I>Reduce</I> section, which lists the time spend in
reducing the per-thread data elements to the storage for non-threaded
computation. These thread timings are taking from the first MPI rank
only and and thus, as the breakdown for MPI tasks can change from MPI
rank to MPI rank, this breakdown can be very different for individual
ranks. Here is an example output for this section:
</P>
<P>Thread timings breakdown (MPI rank 0):
Total threaded time 0.6846 / 90.6%