Demonstrating Hybrid Learning in a Flexible Neuromorphic Hardware System

We present results from a new approach to learning and plasticity in neuromorphic hardware systems: to enable flexibility in implementable learning mechanisms while keeping high efficiency associated with neuromorphic implementations, we combine a general-purpose processor with full-custom analog elements. This processor is operating in parallel with a fully parallel neuromorphic system consisting of an array of synapses connected to analog, continuous time neuron circuits. Novel analog correlation sensor circuits process spike events for each synapse in parallel and in real-time. The processor uses this pre-processing to compute new weights possibly using additional information following its program. Therefore, to a certain extent, learning rules can be defined in software giving a large degree of flexibility. Synapses realize correlation detection geared towards Spike-Timing Dependent Plasticity (STDP) as central computational primitive in the analog domain. Operating at a speed-up factor of 1000 compared to biological time-scale, we measure time-constants from tens to hundreds of micro-seconds. We analyze variability across multiple chips and demonstrate learning using a multiplicative STDP rule. We conclude that the presented approach will enable flexible and efficient learning as a platform for neuroscientific research and technological applications.

[1]  Y. Dan,et al.  Spike timing-dependent plasticity: a Hebbian learning rule. , 2008, Annual review of neuroscience.

[2]  Piotr Dudek,et al.  Compact silicon neuron circuit with spiking and bursting behaviour , 2008, Neural Networks.

[3]  Y. Dan,et al.  Spike Timing-Dependent Plasticity of Neural Circuits , 2004, Neuron.

[4]  Johannes Schemmel,et al.  A wafer-scale neuromorphic hardware system for large-scale neural modeling , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[5]  Steve B. Furber,et al.  A framework for plasticity implementation on the SpiNNaker neural architecture , 2015, Front. Neurosci..

[6]  G. Bi,et al.  Temporal modulation of spike-timing-dependent plasticity , 2022 .

[7]  Wilfred Pinfold,et al.  Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis , 2009, HiPC 2009.

[8]  Trishul M. Chilimbi,et al.  Project Adam: Building an Efficient and Scalable Deep Learning Training System , 2014, OSDI.

[9]  Johannes Schemmel,et al.  Is a 4-Bit Synaptic Weight Resolution Enough? – Constraints on Enabling Spike-Timing Dependent Plasticity in Neuromorphic Hardware , 2012, Front. Neurosci..

[10]  Dake Liu.  Numerical Representation and Finite-Length DSP , 2008 .

[11]  Geoffrey E. Hinton.  Learning multiple layers of representation , 2007, Trends in Cognitive Sciences.

[12]  Wulfram Gerstner,et al.  Limits to high-speed simulations of spiking neural networks using general-purpose computers , 2014, Front. Neuroinform..

[13]  David S. Greenberg,et al.  Population imaging of ongoing neuronal activity in the visual cortex of awake rats , 2008, Nature Neuroscience.

[14]  P. J. Sjöström,et al.  Endocannabinoid-dependent neocortical layer-5 LTD in the absence of postsynaptic spiking. , 2004, Journal of neurophysiology.

[15]  C. Mead,et al.  Neuromorphic analogue VLSI. , 1995, Annual review of neuroscience.

[16]  Karin Strauss,et al.  Accelerating Deep Convolutional Neural Networks Using Specialized Hardware , 2015 .

[17]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[18]  P. J. Sjöström,et al.  Rate, Timing, and Cooperativity Jointly Determine Cortical Synaptic Plasticity , 2001, Neuron.

[19]  Tomoki Fukai,et al.  Supercomputers Ready for Use as Discovery Machines for Neuroscience , 2012, Front. Neuroinform..

[20]  B. Sakmann,et al.  Spiking in primary somatosensory cortex during natural whisking in awake head-restrained rats is cell-type specific , 2009, Proceedings of the National Academy of Sciences.

[21]  James G. King,et al.  Reconstruction and Simulation of Neocortical Microcircuitry , 2015, Cell.

[22]  Johannes J. Letzkus,et al.  Dendritic Synapse Location and Neocortical Spike-Timing-Dependent Plasticity , 2010, Front. Syn. Neurosci..

[23]  G. Bi,et al.  Synaptic Modifications in Cultured Hippocampal Neurons: Dependence on Spike Timing, Synaptic Strength, and Postsynaptic Cell Type , 1998, The Journal of Neuroscience.

[24]  Jürgen Schmidhuber,et al.  Learning Complex, Extended Sequences Using the Principle of History Compression , 1992, Neural Computation.

[25]  Paul E. Hasler,et al.  Floating gate synapses with spike time dependent plasticity , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[26]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[27]  Giacomo Indiveri,et al.  A reconfigurable on-line learning spiking neuromorphic processor comprising 256 neurons and 128K synapses , 2015, Front. Neurosci..

[28]  Tobias C. Potjans,et al.  The Cell-Type Specific Cortical Microcircuit: Relating Structure and Activity in a Full-Scale Spiking Network Model , 2012, Cerebral cortex.

[29]  D. Johnston,et al.  Regulation of Synaptic Efficacy by Coincidence of Postsynaptic APs and EPSPs , 1997 .

[30]  Giacomo Indiveri,et al.  A low-power adaptive integrate-and-fire neuron circuit , 2003, Proceedings of the 2003 International Symposium on Circuits and Systems, 2003. ISCAS '03..

[31]  Wulfram Gerstner,et al.  Phenomenological models of synaptic plasticity based on spike timing , 2008, Biological Cybernetics.

[32]  Karlheinz Meier,et al.  Introducing the Human Brain Project , 2011, FET.

[33]  Jim D. Garside,et al.  Overview of the SpiNNaker System Architecture , 2013, IEEE Transactions on Computers.

[34]  Johannes Schemmel,et al.  Modeling Synaptic Plasticity within Networks of Highly Accelerated I&F Neurons , 2007, 2007 IEEE International Symposium on Circuits and Systems.

[35]  Johannes Schemmel,et al.  Compensating Inhomogeneities of Neuromorphic VLSI Devices Via Short-Term Synaptic Plasticity , 2010, Front. Comput. Neurosci..

[36]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[37]  Tara N. Sainath,et al.  Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research Groups , 2012, IEEE Signal Processing Magazine.

[38]  Arindam Basu,et al.  A Learning-Enabled Neuron Array IC Based Upon Transistor Channel Models of Biological Phenomena , 2013, IEEE Transactions on Biomedical Circuits and Systems.

[39]  Andrew S. Cassidy,et al.  A million spiking-neuron integrated circuit with a scalable communication network and interface , 2014, Science.

[40]  Johannes Schemmel,et al.  Implementing Synaptic Plasticity in a VLSI Spiking Neural Network Model , 2006, The 2006 IEEE International Joint Conference on Neural Network Proceedings.

[41]  Johannes Schemmel,et al.  Reward-based learning under hardware constraints—using a RISC processor embedded in a neuromorphic substrate , 2013, Front. Neurosci..

[42]  Carver A. Mead,et al.  Neuromorphic electronic systems , 1990, Proc. IEEE.

[43]  L. Abbott,et al.  Synaptic plasticity: taming the beast , 2000, Nature Neuroscience.

[44]  Dharmendra S. Modha,et al.  The cat is out of the bag: cortical simulations with 109 neurons, 1013 synapses , 2009, Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis.

[45]  Y. Dan,et al.  Spike timing-dependent plasticity: from synapse to perception. , 2006, Physiological reviews.

[46]  Mostafa Rahimi Azghadi,et al.  Programmable Spike-Timing-Dependent Plasticity Learning Circuits in Neuromorphic VLSI Architectures , 2015, ACM J. Emerg. Technol. Comput. Syst..

[47]  H. Markram,et al.  Regulation of Synaptic Efficacy by Coincidence of Postsynaptic APs and EPSPs , 1997, Science.