Binary Associative Memories as a Benchmark for Spiking Neuromorphic Hardware

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...

Full description

Saved in:
Bibliographic Details
Published in:Frontiers in computational neuroscience Vol. 11; p. 71
Main Authors: Stöckel, Andreas, Jenzen, Christoph, Thies, Michael, Rückert, Ulrich
Format: Journal Article
Language:English
Published: Switzerland Frontiers Research Foundation 22-08-2017
Frontiers Media S.A
Subjects:
associative memory
benchmark
Benchmarks
Computer programs
Energy efficiency
Experiments
Firing pattern
Memory
Neural networks
neuromorphic hardware
Neurons
Neuroscience
Software
spiking neural networks
Workloads
neuromorphic hardware
spiking neural networks
associative memory
benchmark
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary: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.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
Present Address: Andreas Stöckel, Centre for Theoretical Neuroscience, University of Waterloo, Waterloo, ON, Canada
Edited by: Florentin Wörgötter, University of Göttingen, Germany
Reviewed by: Yulia Sandamirskaya, University of Zurich, Switzerland; Markus Diesmann, Forschungszentrum Jülich, Germany
ISSN:1662-5188
1662-5188
DOI:10.3389/fncom.2017.00071