Demonstrating Advantages of Neuromorphic Computation: A Pilot Study

Abstract

Neuromorphic devices represent an attempt to mimic aspects of the brain's architecture and dynamics with the aim of replicating its hallmark functional capabilities in terms of computational power, robust learning and energy efficiency. We employ a single-chip prototype of the BrainScaleS 2 neuromorphic system to implement a proof-of-concept demonstration of reward-modulated spike-timing-dependent plasticity in a spiking network that learns to play a simplified version of the Pong video game by smooth pursuit. This system combines an electronic mixed-signal substrate for emulating neuron and synapse dynamics with an embedded digital processor for on-chip learning, which in this work also serves to simulate the virtual environment and learning agent. The analog emulation of neuronal membrane dynamics enables a 1000-fold acceleration with respect to biological real-time, with the entire chip operating on a power budget of 57 mW. Compared to an equivalent simulation using state-of-the-art software, the on-chip emulation is at least one order of magnitude faster and three orders of magnitude more energy-efficient. We demonstrate how on-chip learning can mitigate the effects of fixed-pattern noise, which is unavoidable in analog substrates, while making use of temporal variability for action exploration. Learning compensates imperfections of the physical substrate, as manifested in neuronal parameter variability, by adapting synaptic weights to match respective excitability of individual neurons.

Artificial Neural Networks and Machine Learning – ICANN 2019: Theoretical Neural Computation

“…We found that a software simulation (NEST v2.14.0) on an Intel processor (i7-4771), when considering only the numerical state propagation, is at least an order of magnitude slower than our neuromorphic emulation (see Fig. 3 and [5]). Besides this, the emulation on our prototype is at least 1000 times more energy-efficient than the software simulation (23 µJ vs. 106 mJ per iteration).…”

Complementary Metal‐Oxide Semiconductor and Memristive Hardware for Neuromorphic Computing

“…Neuromorphic computing was coined by Mead, when he envisioned that while exploiting the similarities between semiconductor physics and biological neural systems, one may develop brain‐inspired computing platforms. Ever since, neuromorphic research has evolved and researchers are implementing various technologies, from conventional semiconductors, as proposed by Mead, to memristive systems, to hybrid CMOS–memristive designs to develop neuro‐mimicking platforms for replicating experimental results observed in biology or for neuro‐inspired platforms used in computing systems …”

Understanding Computational Costs of Cellular-Level Brain Tissue Simulations Through Analytical Performance Models

“…Ultimately we come to the conclusion that while general-purpose computing has, until now, largely been able to deliver high performance, the next generation of brain tissue simulation will be severely limited by hardware bottlenecks. Indeed this trend was already foreshadowed in empirical studies (Jordan et al 2018), and solutions involving hardware accelerators such as General Purpose Graphical Processing Units (GPGPUs) have been proposed (Fidjeland et al 2009;Yavuz et al 2016;Brette and Goodman 2012), while the development of custom brain-like hardware is being actively explored with promising results (van Albada et al 2018;Wunderlich et al 2018).…”