Permalink
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
403 lines (297 sloc) 16.3 KB
==============================================================================
__ __ ___________ _
\ \ / // ___| ___ \ | |
\ V / \ `--.| |_/ / ___ _ __ ___| |__
/ \ `--. \ ___ \/ _ \ '_ \ / __| '_ \
/ /^\ \/\__/ / |_/ / __/ | | | (__| | | |
\/ \/\____/\____/ \___|_| |_|\___|_| |_|
Version 18
==============================================================================
Contact Information
==============================================================================
Organization: Center for Exascale Simulation of Advanced Reactors (CESAR)
Argonne National Laboratory
Development Lead: John Tramm <jtramm@anl.gov>
Ron Rahaman <rahaman@anl.gov>
Amanda Lund <alund@anl.gov>
==============================================================================
What is XSBench?
==============================================================================
XSBench is a mini-app representing a key computational kernel of the
Monte Carlo neutronics application OpenMC.
A full explanation of the theory and purpose of XSBench is provided in
docs/XSBench_Theory.pdf.
==============================================================================
Quick Start Guide
==============================================================================
Download----------------------------------------------------------------------
For the most up-to-date version of XSBench, we recommend that you
download from our git repository. This can be accomplished via
cloning the repository from the command line, or by downloading a zip
from our github page.
Git Repository Clone:
Use the following command to clone XSBench to your machine:
>$ git clone https://github.com/jtramm/XSBench.git
Once cloned, you can update the code to the newest version
using the following command (when in the XSBench directory):
>$ git pull
Git Zip Download:
Simply use the "zip download" option on our webpage at:
https://github.com/jtramm/XSBench
Compilation-------------------------------------------------------------------
To compile XSBench with default settings, use the following
command:
>$ make
Running XSBench---------------------------------------------------------------
To run XSBench with default settings, use the following command:
>$ ./XSBench
For non-default settings, XSBench supports the following command line
options:
Usage: ./XSBench <options>
Options include:
-m <simulation method> Simulation method (history, event)
-t <threads> Number of OpenMP threads to run
-s <size> Size of H-M Benchmark to run (small, large, XL, XXL)
-g <gridpoints> Number of gridpoints per nuclide (overrides -s defaults)
-G <grid type> Grid search type (unionized, nuclide, hash). Defaults to unionized.
-p <particles> Number of particle histories
-l <lookups> Number of Cross-section (XS) lookups per particle history
-h <hash bins> Number of hash bins (only relevant when used with "-G hash")
Default is equivalent to: -s large -l 34 -p 500000 -G unionized
-m <simulation method>
Sets the simulation method, either "history" or "event". These
options represent the history based or event based algorithms
respectively. The default is the history based method. These two
methods represent different methods of parallelizing the Monte
Carlo transport method. In the history based method, the central
mode of parallelism is expressed over particles, which each require
some number of macroscopic cross sections to be executed in series
and in a dependent order. The event based method expresses its
parallelism over a large pool of independent macroscopic cross
section lookups that can be executed in any order without dependence.
They key difference between the two methods is the dependence/independence
of the macroscopic cross section loop.
-t <threads>
Sets the number of OpenMP threads to run. By default, XSBench
will run with 1 thread per hardware core. If the architecture
supports hyperthreading, multiple threads will be run per
core.
If running in MPI mode, this will be the number of threads
per MPI rank.
-s <size>
Sets the size of the Hoogenboom-Martin reactor model. There
are four options: 'small', 'large', 'XL', and 'XXL'. By default,
the 'large' option is selected.
The H-M size corresponds to the number of nuclides present
in the fuel region. The small version has 34 fuel nuclides,
whereas the large version has 321 fuel nuclides. This
significantly slows down the runtime of the program as the
data structures are much larger, and more lookups are required
whenever a lookup occurs in a fuel material. Note that the
program defaults to "Large" if no specification is made.
The additional size options, "XL" and "XXL", do not directly correspond
to any particular physical model. They are similar to the H-M
"large" option, except the number of gridpoints per nuclide
has been increased greatly. This creates an extremely
large energy grid data structure (XL: 120GB, XXL: 252GB), which is
unlikely to fit on a single node, but is useful for experimentation
purposes on novel architectures.
-g <gridpoints>
Sets the number of gridpoints per nuclide. By default, this
value is set to 11,303. This corresponds to the average number
of actual gridpoints per nuclide in the H-M Large model as run
by OpenMC with the actual ACE ENDF cross-section data.
Note that this option will override the number of default grid
-points as set by the '-s' option.
-G <grid type>
Sets the grid search type (unionized, nuclide, hash). Defaults to unionized.
The unionized grid is what is typically used in Monte Carlo codes, as
it offers the fastest speed. However, the increase in speed comes in
a significant increase in memory usage as a union of all the separate
nuclide grids must be formed and stored in memory. The "nuclide" mode
uses only the basic nuclide grid data, with no unionization. This is
slower as a binary search must be performed on every nuclide for each
macroscopic XS lookup, rather than only once when using the unionized
grid.
Finally, the "hash" mode is a newer algorithm now used by
many full Monte Carlo codes which offers speed nearly equivalent
to the unionized energy grid method, but with only a small fraction
of the memory overhead. More details on the hash lookup algorithm
can be found in the CHANGES file for version 16, or in the following
publication:
Forrest B Brown. New hash-based energy lookup algorithm for monte
carlo codes. Trans. Am. Nucl. Soc., 111:659–662, 2014.
http://permalink.lanl.gov/object/tr?what=info:lanl-repo/lareport/LA-UR-14-27037
-p <particles>
Sets the number of particle histories to simulate.
By default, this value is set to 500,000. Users may want to
increase this value if they wish to extend the runtime of
XSBench, perhaps to produce more reliable performance counter
data - as extending the run will decrease the percentage of
runtime spent on initialization. Real MC simulations in a full
application may use up to several billion particles per generation,
so there is great flexibility in this variable.
-l <lookups>
Sets the number of cross-section (XS) lookups to perform per particle.
By default, this value is set to 34, which represents the average
number of XS lookups per particle over the course of its lifetime in
a light water reactor problem. Users should only alter this value if
they are trying to capture the behavior of a different type of reactor
(e.g., one with a fast spectrum), where the number of lookups per
history may be different.
-h <hash bins>
Sets the number of hash bins (only relevant when using the hash
lookup algorithm, as selected with "-G hash"). Default is 10,000.
==============================================================================
Debugging, Optimization & Profiling
==============================================================================
There are also a number of switches that can be set in the makefile.
Here is a sample of the control panel at the top of the makefile:
COMPILER = gnu
OPTIMIZE = yes
DEBUG = no
PROFILE = no
MPI = no
PAPI = no
VEC_INFO = no
VERIFY = no
PAUSE = no
BINARY_DUMP = no
BINARY_READ = no
-> Optimization enables the -O3 optimization flag.
-> Debugging enables the -g flag.
-> Profiling enables the -pg flag. When profiling the code, you may
wish to significantly increase the number of lookups (with the -l
flag) in order to wash out the initialization phase of the code.
-> MPI enables MPI support in the code.
-> The PAPI flag is explained below.
-> VEC_INFO enables some additional information regarding the success or
failure of the compiler's use of vectorization techniques during
compilation.
-> VERIFY enables a verification mode, the details of which are explained below.
-> Binary dump mode writes a binary file containing a randomized data set
of cross sections. This can be used in tandem with the binary read mode
to skip generation of cross section data every time the program is run.
Note that if you create the grid when specifying the -G flag as
"nuclide", data for the unionized energy grid will not be written, and
therefore any subsequent runs using that file in binary read mode must
also use the -G nuclide option. Files generated for the unionized grid
can also be used when running in the nuclide grid mode.
-> Binary read mode reads the binary file created by the binary dump mode
as a (usually) much faster substitution for randomly generating XS
data on-the-fly. This mode is particularly useful if running on
simulators where walltime minimization is extremely critical for
logistical reasons.
==============================================================================
MPI Support
==============================================================================
While XSBench is primarily used to investigate "on node parallelism" issues,
some systems provide power & performance statistics batched in multi-node
configurations. To accommodate this, XSBench provides an MPI mode which
runs the code on all MPI ranks simultaneously. There is no decomposition
across ranks of any kind, and all ranks accomplish the same work. There is
only one point of MPI communication (a reduce) at the end, which aggregates
the timing statistics and averages them across MPI ranks before printing them
out.
MPI support can be enabled with the makefile flag "MPI". If you are not using
the mpicc wrapper on your system, you may need to alter the makefile to
make use of your desired compiler.
==============================================================================
Verification Support
==============================================================================
XSBench has the ability to verify that consistent and correct results are
achieved. This mode is enabled by altering the "VERIFY" setting to 'yes' in
the makefile, i.e.:
VERIFY = yes
Once enabled, the code will generate a hash of the results and display it
with the other data once the code has completed executing. This hash can
then be verified against hashes that other versions or configurations of
the code generate. For instance, running XSBench with 4 threads vs 8 threads
(on a machine that supports that configuration) should generate the
same hash number. Changing the model / run parameters should NOT generate
the same hash number (i.e., increasing the number of particles, number
of gridpoints, etc, will result in different hashes).
Note that the verification mode runs a little slower, due to need to hash
each macroscopic cross section result. Therefore, performance measurements
should generally not be made when verification mode is on. Rather,
verification mode should be used to ensure any changes to the code have not
broken it, and then be disabled before performance metrics are recorded.
==============================================================================
PAPI Performance Counters
==============================================================================
PAPI performance counters is a performance counting library that can
offer information regarding the frequency of specific events (such as
memory loads, cache misses, branch prediction failures, etc) that occur
when the code is executed. XSBench supports use of these performance
counters, although it is left to the user to select the particular
performance counters and locations to instrument.
By default, PAPI is disabled.
To enable PAPI, set in the makefile:
PAPI = yes
Note that you may need to change the relevant library paths for papi to
work (as these are dependent on your machine). The library path can be
specified in the makefile, and the header path is specified in the
XSBench_header.h file.
To select the performance counters you are interested in, open
the file papi.c and alter the events[] array to the events
you would like to count.
==============================================================================
Binary File Support
==============================================================================
The flags:
BINARY_DUMP = no
BINARY_READ = no
Can be set to yes in order to write or read a binary file containing
a randomized XS data set (both nuclide grids and unionized grids). This
feature may be extremely useful for users running on simulators where
walltime minimization is critical for logistical purposes, or for users
who are doing many sequential runs.
Note that identical input parameters (problem size, etc) must be used
when reading and writing a binary file. No runtime checks are made
to validate that the file correctly corresponds to the selected input
parameters.
Also note that if you create the grid when specifying the -G flag as
"nuclide", data for the unionized energy grid will not be written, and
therefore any subsequent runs using that file in binary read mode must
also use the "-G nuclide" option. Files generated for the full unionized grid
can also be used when running in the nuclide grid mode.
==============================================================================
Running on ANL BlueGene/Q (Vesta & Mira)
==============================================================================
Compilation is done using the included makefile, as follows:
>$ make MACHINE=bluegene
Note that the INFO macro in the XSbench_header.h file should be set to
0 when running on BG/Q to remove the run status portions of the output,
which cuts down on unnecessary file I/O, i.e.:
#define INFO 0
Also, note that you may need to add the following line to your .soft
file in order to use the mpicc compiler wrapper:
+mpiwrapper-gcc
Then, be sure to use the "resoft" command to update your software, i.e.,:
>$ resoft
When running in c16 mode, the maximum number of gridpoints per nuclide
is 900 (when running in "Large" mode). More points will cause the 1GB
memory limit to be broken.
A basic test run on 1 node can be achieved (assuming you have an allocation)
using the makefile and the following command:
>$ make bgqrun
Further information on queuing can be found at:
https://www.alcf.anl.gov/resource-guides/vesta-queuing
==============================================================================
Citing XSBench
==============================================================================
Papers citing the XSBench program in general should refer to:
J. R. Tramm, A. R. Siegel, T. Islam, and M. Schulz, “XSBench - The
Development and Verification of a Performance Abstraction for Monte
Carlo Reactor Analysis,” presented at PHYSOR 2014 - The Role
of Reactor Physics toward a Sustainable Future, Kyoto.
A PDF of this paper can be accessed directly at this link:
http://www.mcs.anl.gov/papers/P5064-0114.pdf
Bibtex Entry:
@inproceedings{Tramm:wy,
author = {Tramm, John R and Siegel, Andrew R and Islam, Tanzima and Schulz, Martin},
title = {{XSBench - The Development and Verification of a Performance Abstraction for Monte Carlo Reactor Analysis}},
booktitle = {PHYSOR 2014 - The Role of Reactor Physics toward a Sustainable Future},
address = {Kyoto}
}
==============================================================================