Scalable network emulation on analog neuromorphic hardware
We present a novel software feature for the BrainScaleS-2 accelerated neuromorphic platform that facilitates the partitioned emulation of large-scale spiking neural networks. This approach is well suited for deep spiking neural networks and allows for sequential model emulation on undersized neuromo...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-02-01
|
| Series: | Frontiers in Neuroscience |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fnins.2024.1523331/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832539906909405184 |
|---|---|
| author | Elias Arnold Philipp Spilger Jan V. Straub Eric Müller Dominik Dold Gabriele Meoni Johannes Schemmel |
| author_facet | Elias Arnold Philipp Spilger Jan V. Straub Eric Müller Dominik Dold Gabriele Meoni Johannes Schemmel |
| author_sort | Elias Arnold |
| collection | DOAJ |
| description | We present a novel software feature for the BrainScaleS-2 accelerated neuromorphic platform that facilitates the partitioned emulation of large-scale spiking neural networks. This approach is well suited for deep spiking neural networks and allows for sequential model emulation on undersized neuromorphic resources if the largest recurrent subnetwork and the required neuron fan-in fit on the substrate. We demonstrate the training of two deep spiking neural network models—using the MNIST and EuroSAT datasets—that exceed the physical size constraints of a single-chip BrainScaleS-2 system. The ability to emulate and train networks larger than the substrate provides a pathway for accurate performance evaluation in planned or scaled systems, ultimately advancing the development and understanding of large-scale models and neuromorphic computing architectures. |
| format | Article |
| id | doaj-art-77add19c44f5426fad96df2a5e8aa027 |
| institution | Kabale University |
| issn | 1662-453X |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroscience |
| spelling | doaj-art-77add19c44f5426fad96df2a5e8aa0272025-02-05T07:32:46ZengFrontiers Media S.A.Frontiers in Neuroscience1662-453X2025-02-011810.3389/fnins.2024.15233311523331Scalable network emulation on analog neuromorphic hardwareElias Arnold0Philipp Spilger1Jan V. Straub2Eric Müller3Dominik Dold4Gabriele Meoni5Johannes Schemmel6European Institute for Neuromorphic Computing, Kirchhoff Institute for Physics, Heidelberg University, Heidelberg, GermanyEuropean Institute for Neuromorphic Computing, Kirchhoff Institute for Physics, Heidelberg University, Heidelberg, GermanyEuropean Institute for Neuromorphic Computing, Kirchhoff Institute for Physics, Heidelberg University, Heidelberg, GermanyEuropean Institute for Neuromorphic Computing, Kirchhoff Institute for Physics, Heidelberg University, Heidelberg, GermanyAdvanced Concepts Team, European Space Research and Technology Centre, European Space Agency, Noordwijk, NetherlandsFaculty of Aerospace Engineering, Delft University of Technology, Delft, NetherlandsEuropean Institute for Neuromorphic Computing, Kirchhoff Institute for Physics, Heidelberg University, Heidelberg, GermanyWe present a novel software feature for the BrainScaleS-2 accelerated neuromorphic platform that facilitates the partitioned emulation of large-scale spiking neural networks. This approach is well suited for deep spiking neural networks and allows for sequential model emulation on undersized neuromorphic resources if the largest recurrent subnetwork and the required neuron fan-in fit on the substrate. We demonstrate the training of two deep spiking neural network models—using the MNIST and EuroSAT datasets—that exceed the physical size constraints of a single-chip BrainScaleS-2 system. The ability to emulate and train networks larger than the substrate provides a pathway for accurate performance evaluation in planned or scaled systems, ultimately advancing the development and understanding of large-scale models and neuromorphic computing architectures.https://www.frontiersin.org/articles/10.3389/fnins.2024.1523331/fullmodelingneuromorphicspiking neural networksvirtualizationaccelerator abstraction |
| spellingShingle | Elias Arnold Philipp Spilger Jan V. Straub Eric Müller Dominik Dold Gabriele Meoni Johannes Schemmel Scalable network emulation on analog neuromorphic hardware Frontiers in Neuroscience modeling neuromorphic spiking neural networks virtualization accelerator abstraction |
| title | Scalable network emulation on analog neuromorphic hardware |
| title_full | Scalable network emulation on analog neuromorphic hardware |
| title_fullStr | Scalable network emulation on analog neuromorphic hardware |
| title_full_unstemmed | Scalable network emulation on analog neuromorphic hardware |
| title_short | Scalable network emulation on analog neuromorphic hardware |
| title_sort | scalable network emulation on analog neuromorphic hardware |
| topic | modeling neuromorphic spiking neural networks virtualization accelerator abstraction |
| url | https://www.frontiersin.org/articles/10.3389/fnins.2024.1523331/full |
| work_keys_str_mv | AT eliasarnold scalablenetworkemulationonanalogneuromorphichardware AT philippspilger scalablenetworkemulationonanalogneuromorphichardware AT janvstraub scalablenetworkemulationonanalogneuromorphichardware AT ericmuller scalablenetworkemulationonanalogneuromorphichardware AT dominikdold scalablenetworkemulationonanalogneuromorphichardware AT gabrielemeoni scalablenetworkemulationonanalogneuromorphichardware AT johannesschemmel scalablenetworkemulationonanalogneuromorphichardware |