Limits to high-speed simulations of spiking neural networks using general-purpose computers
To understand how the central nervous system performs computations using recurrent neuronal circuitry, simulations have become an indispensable tool for theoretical neuroscience. To study neuronal circuits and their ability to self-organize, increasing attention has been directed towards synaptic pl...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2014-09-01
|
| Series: | Frontiers in Neuroinformatics |
| Subjects: | |
| Online Access: | http://journal.frontiersin.org/Journal/10.3389/fninf.2014.00076/full |
| _version_ | 1818431395068903424 |
|---|---|
| author | Friedemann eZenke Friedemann eZenke Wulfram eGerstner Wulfram eGerstner |
| author_facet | Friedemann eZenke Friedemann eZenke Wulfram eGerstner Wulfram eGerstner |
| author_sort | Friedemann eZenke |
| collection | DOAJ |
| description | To understand how the central nervous system performs computations using recurrent neuronal circuitry, simulations have become an indispensable tool for theoretical neuroscience. To study neuronal circuits and their ability to self-organize, increasing attention has been directed towards synaptic plasticity. In particular spike-timing-dependent plasticity (STDP) creates specific demands for simulations of spiking neural networks. On the one hand a high temporal resolution is required to capture the millisecond timescale of typical STDP windows. On the other hand network simulations have to evolve over hours up to days, to capture the timescale of long-term plasticity. To do this efficiently, fast simulation speed is the crucial ingredient rather than large neuron numbers. Using different medium-sized network models consisting of several thousands of neurons and off-the-shelf hardware, we compare the simulation speed of the simulators: Brian, NEST and Neuron as well as our own simulator Auryn. Our results show that real-time simulations of different plastic network models are possible in parallel simulations in which numerical precision is not a primary concern. Even so, the speed-up margin of parallelism is limited and boosting simulation speeds beyond one tenth of real-time is difficult. By profiling simulation code we show that the run times of typical plastic network simulations encounter a hard boundary. This limit is partly due to latencies in the inter-process communications and thus cannot be overcome by increased parallelism. Overall these results show that to study plasticity in medium-sized spiking neural networks, adequate simulation tools are readily available which run efficiently on small clusters. However, to run simulations substantially faster than real-time, special hardware is a prerequisite. |
| first_indexed | 2024-12-14T15:48:37Z |
| format | Article |
| id | doaj.art-75dd3c5367344f21b7ae233d22da6a66 |
| institution | Directory Open Access Journal |
| issn | 1662-5196 |
| language | English |
| last_indexed | 2024-12-14T15:48:37Z |
| publishDate | 2014-09-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroinformatics |
| spelling | doaj.art-75dd3c5367344f21b7ae233d22da6a662022-12-21T22:55:25ZengFrontiers Media S.A.Frontiers in Neuroinformatics1662-51962014-09-01810.3389/fninf.2014.0007689449Limits to high-speed simulations of spiking neural networks using general-purpose computersFriedemann eZenke0Friedemann eZenke1Wulfram eGerstner2Wulfram eGerstner3École polytechnique fédérale de LausanneÉcole polytechnique fédérale de LausanneÉcole polytechnique fédérale de LausanneÉcole polytechnique fédérale de LausanneTo understand how the central nervous system performs computations using recurrent neuronal circuitry, simulations have become an indispensable tool for theoretical neuroscience. To study neuronal circuits and their ability to self-organize, increasing attention has been directed towards synaptic plasticity. In particular spike-timing-dependent plasticity (STDP) creates specific demands for simulations of spiking neural networks. On the one hand a high temporal resolution is required to capture the millisecond timescale of typical STDP windows. On the other hand network simulations have to evolve over hours up to days, to capture the timescale of long-term plasticity. To do this efficiently, fast simulation speed is the crucial ingredient rather than large neuron numbers. Using different medium-sized network models consisting of several thousands of neurons and off-the-shelf hardware, we compare the simulation speed of the simulators: Brian, NEST and Neuron as well as our own simulator Auryn. Our results show that real-time simulations of different plastic network models are possible in parallel simulations in which numerical precision is not a primary concern. Even so, the speed-up margin of parallelism is limited and boosting simulation speeds beyond one tenth of real-time is difficult. By profiling simulation code we show that the run times of typical plastic network simulations encounter a hard boundary. This limit is partly due to latencies in the inter-process communications and thus cannot be overcome by increased parallelism. Overall these results show that to study plasticity in medium-sized spiking neural networks, adequate simulation tools are readily available which run efficiently on small clusters. However, to run simulations substantially faster than real-time, special hardware is a prerequisite.http://journal.frontiersin.org/Journal/10.3389/fninf.2014.00076/fullSTDPsynaptic plasticitycomputational neuroscienceParallel ComputingRecurrent networksspiking neural networks |
| spellingShingle | Friedemann eZenke Friedemann eZenke Wulfram eGerstner Wulfram eGerstner Limits to high-speed simulations of spiking neural networks using general-purpose computers Frontiers in Neuroinformatics STDP synaptic plasticity computational neuroscience Parallel Computing Recurrent networks spiking neural networks |
| title | Limits to high-speed simulations of spiking neural networks using general-purpose computers |
| title_full | Limits to high-speed simulations of spiking neural networks using general-purpose computers |
| title_fullStr | Limits to high-speed simulations of spiking neural networks using general-purpose computers |
| title_full_unstemmed | Limits to high-speed simulations of spiking neural networks using general-purpose computers |
| title_short | Limits to high-speed simulations of spiking neural networks using general-purpose computers |
| title_sort | limits to high speed simulations of spiking neural networks using general purpose computers |
| topic | STDP synaptic plasticity computational neuroscience Parallel Computing Recurrent networks spiking neural networks |
| url | http://journal.frontiersin.org/Journal/10.3389/fninf.2014.00076/full |
| work_keys_str_mv | AT friedemannezenke limitstohighspeedsimulationsofspikingneuralnetworksusinggeneralpurposecomputers AT friedemannezenke limitstohighspeedsimulationsofspikingneuralnetworksusinggeneralpurposecomputers AT wulframegerstner limitstohighspeedsimulationsofspikingneuralnetworksusinggeneralpurposecomputers AT wulframegerstner limitstohighspeedsimulationsofspikingneuralnetworksusinggeneralpurposecomputers |