Merge branch 'master' into model-doc-order

Author: jessica-mitchell

doc/htmldoc/benchmark_results.rst

@@ -0,0 +1,106 @@
+.. _nest_benchmark_results:
+
+NEST performance benchmarks
+===========================
+
+
+NEST performance is continuously monitored and improved across various network sizes.
+Here we show benchmarking results for NEST version 3.8 on Jureca-DC.
+
+
+Strong scaling experiment of the Microcircuit model [1]_
+---------------------------------------------------------
+
+.. grid:: 1 1 1 1
+
+   .. grid-item::
+       :columns: 10
+       :class: sd-align-major-center
+
+       .. image:: /static/img/mc_benchmark.png
+
+.. grid:: 1 1 1 1
+
+   .. grid-item::
+       :columns: 10
+       :class: sd-align-minor-center
+
+
+       * The model has ~80 000 neurons and ~300 million synapses
+       * Increasing number of computing resources decrease simulation time
+       * The model runs faster than real time
+
+
+
+
+Strong scaling experiment of the Multi-area-model [2]_
+-------------------------------------------------------
+
+.. grid:: 1 1 1 1
+
+ .. grid-item::
+       :class: sd-align-major-center
+       :columns: 10
+
+       .. image:: /static/img/mam_benchmark.png
+
+
+.. grid:: 1 1 1 1
+
+ .. grid-item::
+       :columns: 10
+       :class: sd-align-minor-center
+
+       * The model has ~4.1 million neurons and ~24 billion synapses
+       * Steady decrease of run time with additional compute resources
+
+
+
+
+Weak scaling experiment of the HPC benchmark model [3]_
+--------------------------------------------------------
+
+.. grid:: 1 1 1 1
+
+   .. grid-item::
+       :columns: 10
+       :class: sd-align-major-center
+
+       .. image:: /static/img/hpc_benchmark.png
+
+
+.. grid:: 1 1 1 1
+
+   .. grid-item::
+       :columns: 10
+       :class: sd-align-minor-center
+
+
+       * The size of network scales proportionally with the computational resources used
+       * Largest network size in this diagram: ~5.8 million neurons and ~65 billion synapses
+       * The figure shows that NEST can handle massive networks and simulate them efficiently
+
+
+.. seealso::
+
+   Example networks:
+
+   * :doc:`/auto_examples/Potjans_2014/index`
+   * `Multi-area model <https://inm-6.github.io/multi-area-model/>`_
+   * :doc:`/auto_examples/hpc_benchmark`
+
+References
+----------
+
+.. [1]  Potjans TC. and Diesmann M. 2014. The cell-type specific cortical
+        microcircuit: relating structure and activity in a full-scale spiking
+        network model. Cerebral Cortex. 24(3):785–806. DOI: `10.1093/cercor/bhs358 <https://doi.org/10.1093/cercor/bhs358>`__.
+
+
+.. [2] Schmidt M, Bakker R, Hilgetag CC, Diesmann M and van Albada SJ. 2018. Multi-scale
+       account of the network structure of macaque visual cortex. Brain Structure
+       and Function. 223: 1409 https://doi.org/10.1007/s00429-017-1554-4
+
+.. [3] Jordan J, Ippen T, Helias M, Kitayama I, Sato M, Igarashi J, Diesmann M, Kunkel S. 2018.
+       Extremely scalable spiking neuronal network simulation code: From laptops to exacale computers.
+       Frontiers in Neuroinformatics. 12. https://www.frontiersin.org/journals/neuroinformatics/articles/10.3389/fninf.2018.00002

doc/htmldoc/hpc/benchmarking.rst

@@ -7,6 +7,12 @@ Benchmarking NEST
 
    When compiling NEST to perform benchmarks, see :ref:`our cmake options <performance_cmake>` for improved performance and energy saving.
 
+NEST performance
+~~~~~~~~~~~~~~~~
+
+We regulary test NEST performance at various network sizes.
+For results of these experiments see our :ref:`nest_benchmark_results` page.
+
 beNNch
 ~~~~~~
 

doc/htmldoc/hpc/optimizing_nest.rst

@@ -18,7 +18,7 @@ hardware and software of the HPC system you are using for improving the overall
 performance.
 
 
-Although there will be some variation between scripts and HPC systems, in general, we recommend that 
+Although there will be some variation between scripts and HPC systems, in general, we recommend that
 
  * resources are fully utilized (e.g., all available cores are used).
  * one thread pinned to one core (no simultaneous multithreading)
@@ -59,7 +59,11 @@ Although there will be some variation between scripts and HPC systems, in genera
 
 .. seealso::
 
-   See our other guides on 
+   See our other guides on
 
    * :ref:`parallel_computing`
    * :ref:`benchmark`
+
+   or see our performance results:
+
+   * :ref:`nest_benchmark_results`

doc/htmldoc/index.rst

@@ -123,6 +123,7 @@ How NEST works --- The Big Picture
    Models <models/index>
    Python API <ref_material/pynest_api/index>
    ref_material/glossary
+   benchmark_results
    Cite NEST <citing-nest>
    License <license>
 

doc/htmldoc/static/img/hpc_benchmark.png

@@ -2239,6 +2239,7 @@ special prerequisites.
 SeeAlso: exit_test_gracefully
 */
 
+/skip_if_without_boost { statusdict/have_boost :: not { /skipped_no_boost exit_test_gracefully } if } def
 /skip_if_no_mpi { statusdict/is_mpi :: not { /skipped_no_mpi exit_test_gracefully } if } def
 /skip_if_have_mpi { statusdict/is_mpi :: { /skipped_have_mpi exit_test_gracefully } if } def
 /skip_if_not_threaded { is_threaded not { /skipped_no_threading exit_test_gracefully } if } def

doc/htmldoc/static/img/mam_benchmark.png

@@ -23,6 +23,7 @@
 #include "iaf_bw_2001.h"
 
 #ifdef HAVE_GSL
+#ifdef HAVE_BOOST
 
 // Includes from libnestutil:
 #include "dict_util.h"
@@ -521,4 +522,5 @@ nest::iaf_bw_2001::handle( CurrentEvent& e )
     e.get_rel_delivery_steps( kernel().simulation_manager.get_slice_origin() ), e.get_weight() * e.get_current() );
 }
 
+#endif // HAVE_BOOST
 #endif // HAVE_GSL

doc/htmldoc/static/img/mc_benchmark.png

@@ -27,6 +27,7 @@
 #include "config.h"
 
 #ifdef HAVE_GSL
+#ifdef HAVE_BOOST
 
 // C includes:
 #include <gsl/gsl_errno.h>
@@ -516,5 +517,6 @@ iaf_bw_2001::set_status( const DictionaryDatum& d )
 };
 } // namespace
 
+#endif // HAVE_BOOST
 #endif // HAVE_GSL
 #endif // IAF_BW_2001