Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware
Large-scale neuromorphic hardware platforms, specialized computer systems for energy efficient simulation of spiking neural networks, are being developed around the world, for example as part of the European Human Brain Project (HBP). Due to conceptual differences, a universal performance analysis o...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2017-08-01
|
| Series: | Frontiers in Computational Neuroscience |
| Subjects: | |
| Online Access: | http://journal.frontiersin.org/article/10.3389/fncom.2017.00071/full |
| _version_ | 1830441554447171584 |
|---|---|
| author | Andreas Stöckel Christoph Jenzen Michael Thies Ulrich Rückert |
| author_facet | Andreas Stöckel Christoph Jenzen Michael Thies Ulrich Rückert |
| author_sort | Andreas Stöckel |
| collection | DOAJ |
| description | Large-scale neuromorphic hardware platforms, specialized computer systems for energy efficient simulation of spiking neural networks, are being developed around the world, for example as part of the European Human Brain Project (HBP). Due to conceptual differences, a universal performance analysis of these systems in terms of runtime, accuracy and energy efficiency is non-trivial, yet indispensable for further hard- and software development. In this paper we describe a scalable benchmark based on a spiking neural network implementation of the binary neural associative memory. We treat neuromorphic hardware and software simulators as black-boxes and execute exactly the same network description across all devices. Experiments on the HBP platforms under varying configurations of the associative memory show that the presented method allows to test the quality of the neuron model implementation, and to explain significant deviations from the expected reference output. |
| first_indexed | 2024-12-21T05:23:44Z |
| format | Article |
| id | doaj.art-109ebbb7601c40339fdad9ff5e65ffe1 |
| institution | Directory Open Access Journal |
| issn | 1662-5188 |
| language | English |
| last_indexed | 2024-12-21T05:23:44Z |
| publishDate | 2017-08-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Computational Neuroscience |
| spelling | doaj.art-109ebbb7601c40339fdad9ff5e65ffe12022-12-21T19:14:44ZengFrontiers Media S.A.Frontiers in Computational Neuroscience1662-51882017-08-011110.3389/fncom.2017.00071244636Binary Associative Memories as a Benchmark for Spiking Neuromorphic HardwareAndreas StöckelChristoph JenzenMichael ThiesUlrich RückertLarge-scale neuromorphic hardware platforms, specialized computer systems for energy efficient simulation of spiking neural networks, are being developed around the world, for example as part of the European Human Brain Project (HBP). Due to conceptual differences, a universal performance analysis of these systems in terms of runtime, accuracy and energy efficiency is non-trivial, yet indispensable for further hard- and software development. In this paper we describe a scalable benchmark based on a spiking neural network implementation of the binary neural associative memory. We treat neuromorphic hardware and software simulators as black-boxes and execute exactly the same network description across all devices. Experiments on the HBP platforms under varying configurations of the associative memory show that the presented method allows to test the quality of the neuron model implementation, and to explain significant deviations from the expected reference output.http://journal.frontiersin.org/article/10.3389/fncom.2017.00071/fullneuromorphic hardwarespiking neural networksbenchmarkassociative memory |
| spellingShingle | Andreas Stöckel Christoph Jenzen Michael Thies Ulrich Rückert Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware Frontiers in Computational Neuroscience neuromorphic hardware spiking neural networks benchmark associative memory |
| title | Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware |
| title_full | Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware |
| title_fullStr | Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware |
| title_full_unstemmed | Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware |
| title_short | Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware |
| title_sort | binary associative memories as a benchmark for spiking neuromorphic hardware |
| topic | neuromorphic hardware spiking neural networks benchmark associative memory |
| url | http://journal.frontiersin.org/article/10.3389/fncom.2017.00071/full |
| work_keys_str_mv | AT andreasstockel binaryassociativememoriesasabenchmarkforspikingneuromorphichardware AT christophjenzen binaryassociativememoriesasabenchmarkforspikingneuromorphichardware AT michaelthies binaryassociativememoriesasabenchmarkforspikingneuromorphichardware AT ulrichruckert binaryassociativememoriesasabenchmarkforspikingneuromorphichardware |