Auryn simulator

Simulator for spiking neural networks with synaptic plasticity

User Tools

Site Tools


start

This is an old revision of the document!


Auryn spiking neural network simulator

Auryn is a simulator for spiking neural networks with plastic synapses. More specifically, it is free and open source software optimized to build and simulate medium sized spiking neural networks with plasticity (e.g. Vogels et al., Science, 2011; Zenke et al., Nature Communications, 2015; and more). It is complementary to existing neural simulators such as NEST and Brian, of which many successful concepts it adopts. Its main focus lies on performance to allow simulating plastic networks with sub-millisecond temporal resolution over extended periods of time (hours to days) efficiently. Auryn is written from scratch in C++ and can be run in parallel on clusters using MPI. Auryn also runs on Intel's Xeon Phi architecture and has experimental GPU support.

Guiding principles

  • Modularity. A network model is defined as a collection of objects (e.g. a group of neurons) and the interactions between them (e.g. sparse synaptic connectivity). Additionally, the user defines devices such as monitors to read out activity in the network simulation. The Auryn kernel keeps track of all the simulation objects and runs the simulation (also transparently on a cluster if desired).
  • Performance. Each network object is kept short and simple and optimized for the task at hand. Data is stored in customized vector classes and when possible Auryn uses vectorization such as SSE registers to speed up operations on them.
  • Easily extensible. New modules are easy to add (as new neuron or synapse types), and existing modules provide good examples as to how.

This wiki aims at providing examples and documentation for Auryn users and developers. To get started take a look at quick start and check out the Examples pages. Here you will also find most of the simulation code of published work. The best place for questions is the support forum.

To cite Auryn: Zenke, F. and Gerstner, W., 2014. Limits to high-speed simulations of spiking neural networks using general-purpose computers. Front Neuroinform 8, 76. doi: 10.3389/fninf.2014.00076

Copyright and License Auryn is written and maintained by Friedemann Zenke (2014-2016). Auryn is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. — Auryn is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. — You should have received a copy of the GNU General Public License along with Auryn. If not, see <http://www.gnu.org/licenses/>.

The Auryn logo Copyright 2013 Koshika Yadava. The term Auryn refers to an artifact in “The Never-ending Story”, a novel by Michael Ende.

Contribute If you want to have write access to this Wiki, let me know. To learn how to write and modify entries please check out the syntax section. If you are looking for a good place to start, you find a list with wanted articles.

Site Tracking/Analytics Only if your browser privacy settings allow tracking, this page tracks visits. This helps me improve the most relevant topics. You can opt out explicitly from tracking using this link.

Funding During Auryn's development Friedemann received support of the European Community's Seventh Framework Program under grant agreement no. 237955 (FACETS-ITN), 269921 (BrainScales) and the European Research Council under grant agreement no. 268689 (MultiRules). Moreover, Friedemann was supported by the SNSF (Swiss National Science Foundation).

start.1472772255.txt.gz · Last modified: 2016/09/01 23:24 by zenke