A wafer-scale neuromorphic hardware system for large-scale neural modeling

Schemmel, Johannes; Briiderle, Daniel; Griibl, Andreas; Hock, Matthias; Meier, K.; Millner, Sebastian

doi:10.1109/iscas.2010.5536970

Cited by 578 publications

(509 citation statements)

References 11 publications

Supporting

Mentioning

494

Contrasting

Unclassified

Order By: Relevance

“…Finally, only a small number of connections between the transition neurons and the state populations need to be specified or learned to achieve a desired functionality. This property can be useful for framing the design of learning algorithms and for reducing the number of on-chip plastic synapses required to implement autonomous learning of behaviors (7,49,50). The complexity of a learning problem is proportional to the dimensionality of the to-be-learned parameters (51).…”

Section: Discussionmentioning

confidence: 99%

“…So far, research emphasis has been on developing largescale neural-like electronic systems (6)(7)(8). However, the concepts and methods for installing the dynamics necessary to express cognitive behaviors on these substrates are relatively poorly developed, mainly because the deep question of how biological brains install cognition on their neural networks remains open.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Synthesizing cognition in neuromorphic electronic systems

Neftci

Binas

Rutishauser

et al. 2013

Proc. Natl. Acad. Sci. U.S.A.

127

122

View full text Add to dashboard Cite

The quest to implement intelligent processing in electronic neuromorphic systems lacks methods for achieving reliable behavioral dynamics on substrates of inherently imprecise and noisy neurons. Here we report a solution to this problem that involves first mapping an unreliable hardware layer of spiking silicon neurons into an abstract computational layer composed of generic reliable subnetworks of model neurons and then composing the target behavioral dynamics as a "soft state machine" running on these reliable subnets. In the first step, the neural networks of the abstract layer are realized on the hardware substrate by mapping the neuron circuit bias voltages to the model parameters. This mapping is obtained by an automatic method in which the electronic circuit biases are calibrated against the model parameters by a series of population activity measurements. The abstract computational layer is formed by configuring neural networks as generic soft winner-take-all subnetworks that provide reliable processing by virtue of their active gain, signal restoration, and multistability. The necessary states and transitions of the desired high-level behavior are then easily embedded in the computational layer by introducing only sparse connections between some neurons of the various subnets. We demonstrate this synthesis method for a neuromorphic sensory agent that performs real-time context-dependent classification of motion patterns observed by a silicon retina.decision making | sensorimotor | working memory | analog very large-scale integration | artificial neural systems U nlike digital simulations, in which the dynamics of neuronal models are encoded and calculated on general purpose digital hardware, "neuromorphic" emulations express the dynamics of the neural systems directly on an analogous physical substrate (1). Digital simulations have the advantage that they can be exactly and reliably programmed using numerical operations of very high precision. However, they suffer the disadvantage that they are cast on abstract binary electronic circuits whose operation is entirely divorced from the physical processes being simulated. Consequently, such simulations do not readily advance our understanding of how biological neural systems are able to attain their extraordinary physical performance, using only large numbers of apparently unreliable, slow, and imprecise neural components, a problem first recognized by von Neumann more than a half century ago (2). Furthermore, the reliability of digital systems comes at high cost of the circuit complexity necessary for the orchestration of communication and processing, an overhead that declares itself also in the costs of system construction and power dissipation (3).The alternative, neuromorphic, approach to information processing strives to capture in complementary metal-oxide semiconductor (CMOS) very large-scale integration (VLSI) electronic technology the more distributed, asynchronous, and limited precision nature of biological intelligent systems (1, 4).†, ‡ Research...

show abstract

Section: Discussionmentioning

confidence: 99%

mentioning

confidence: 99%

Synthesizing cognition in neuromorphic electronic systems

Neftci

Binas

Rutishauser

et al. 2013

Proc. Natl. Acad. Sci. U.S.A.

127

122

View full text Add to dashboard Cite

show abstract

“…Traditionally, electronic techniques have been mainly used for neuromorphic systems yielding already platforms such as Neurogrid at Stanford [2], TrueNorth at IBM [3], HICANN at the University of Heidelberg [4] and the University of Manchester s neuromorphic chip Electronic techniques however face important challenges, e.g. limited bandwidth, large multicasting and communication issues, which ultimately impose performance limits [6].…”

mentioning

confidence: 99%

Controlled inhibition of spiking dynamics in VCSELs for neuromorphic photonics: theory and experiments

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Since Spikey also supports on-chip short-term plasticity (STP) and spike-timing dependent synaptic plasticity (STDP), it is predestined for high-performance spike-based computing. Spikey is the predecessor of a large, wafer-scale system currently developed in Heidelberg [29], which is aimed at large-scale neuromorphic computing for accelerated biological simulation and data mining. Figure 1 shows a schematic overview of the Spikey system.…”

Section: Methodological Background the Spikey Neuromorphic Hardware Smentioning

confidence: 99%

“…Several platforms for neuromorphic computing have emerged, starting with the pioneering work of Carver Mead [23], up to rather recent developments like e.g. SpiNNaker [24], NeuroGrid [25], ROLLS [26], IBM's TrueNorth [27], or the systems developed at the University of Heidelberg [28], [29] (see [30] for a review). Aside from testing principles of neural computation [28], [31], [32], initial applications of neuromorphic hardware have been demonstrated for generic pattern recognition [27], [33].…”

Section: Introductionmentioning

confidence: 99%

Predicting voluntary movements from motor cortical activity with neuromorphic hardware

Lungu¹,

Riehle²,

Nawrot³

et al. 2017

IBM J. Res. & Dev.

View full text Add to dashboard Cite

Neurons in mammalian motor cortex encode physical parameters of voluntary movements during planning and execution of a motor task. Brain-machine interfaces can decode limb movements from the activity of these neurons in real time. The future goal is to control prosthetic devices in severely paralyzed patients or to restore communication if the ability to speak or make gestures is lost. Here, we implemented a spiking neural network that decodes movement intentions from the activity of individual neurons recorded in the motor cortex of a monkey. The network runs on neuromorphic hardware and performs its computations in a purely spike-based fashion. It incorporates an insect-brain-inspired, three-layer architecture with 176 neurons. Cortical signals are filtered using lateral inhibition, and the network is trained in a supervised fashion to predict two opposing directions of the monkey's arm reaching movement before the movement is carried out. Our network operates on the actual spikes that have been emitted by motor cortical neurons, without the need to construct intermediate non-spiking representations. Using a pseudo-population of 12 manually-selected neurons, it reliably predicts the movement direction with an accuracy of 89.32 % on unseen data after only 100 training trials. Our results provide a proof of concept for the first-time use of a neuromorphic device for decoding movement intentions.

show abstract

A wafer-scale neuromorphic hardware system for large-scale neural modeling

Cited by 578 publications

References 11 publications

Synthesizing cognition in neuromorphic electronic systems

Synthesizing cognition in neuromorphic electronic systems

Controlled inhibition of spiking dynamics in VCSELs for neuromorphic photonics: theory and experiments

Predicting voluntary movements from motor cortical activity with neuromorphic hardware

Contact Info

Product

Resources

About