Since the beginning of the computer era, the human brain inspires scientists as an alternative approach to artificial machine intelligence. Whereas artificial neural networks are highly abstracted models of brain-inspired computing, so called neuromorphic systems emulate the biological paragon in more detail. The main difference of this approach is the use of the more bio-inspired neuron models like the (leaky) integrate-and-fire neuron or more complex neuron models. These neuron models communicate via spikes and are computationally more complex. This chapter gives an update of the latest developments in brain-inspired neuromorphic systems. New features of the second generation of the SpiNNaker and the BrainScaleS systems from the European Human Brain Project will be presented. In addition, new approaches like the Intel Loihi system are introduced. The scalability and performance of selected systems will be compared theoretically and based on benchmark tests.
CITATION STYLE
Rueckert, U. (2020). Update on Brain-Inspired Systems. In Frontiers Collection (Vol. Part F1076, pp. 387–403). Springer VS. https://doi.org/10.1007/978-3-030-18338-7_22
Mendeley helps you to discover research relevant for your work.