Systems
The field is defined as much by the machines built as by theory. These are the systems most often cited; all figures are as published by their builders.
| System | Who and when | What is distinctive |
|---|---|---|
| Silicon retina | Carver Mead and Misha Mahowald, Caltech; first published 1988, widely known from 1991 | Among the first neuromorphic sensors: analog circuits that mimic the first layers of the eye |
| Dynamic vision sensor | Lichtsteiner, Posch and Delbruck, Zurich; first shown 2006, full paper 2008 | A camera whose pixels report only changes in brightness, each independently, with microsecond timing |
| Neurogrid | Stanford University, 2014 | A board of sixteen mixed analog–digital chips simulating a million neurons in real time |
| TrueNorth | IBM, 2014 | One million digital spiking neurons and 256 million synapses on one chip, at about 70 milliwatts |
| SpiNNaker | University of Manchester; million-core machine switched on in 2018 | Very many small conventional processor cores joined by a network built for spikes |
| BrainScaleS | Heidelberg University | Analog neuron circuits that run far faster than biological real time |
| Loihi and Loihi 2 | Intel, 2017 and 2021 | Asynchronous digital chips with programmable on-chip learning |
| Tianjic | Tsinghua University, 2019 | A hybrid chip running spiking and conventional neural networks together |
| Hala Point | Intel, installed at Sandia National Laboratories, 2024 | A system of 1.15 billion neurons, the largest announced at the time |
| Memristive crossbar arrays | Many laboratories, since 2008 | Grids of resistive memory elements that multiply and sum in place, in the analog domain |
Not the same as a neural network accelerator
The chips that train and run today’s large neural networks are fast, dense, clocked arithmetic engines; they borrow the brain’s wiring diagram loosely and none of its signalling. Neuromorphic hardware borrows the signalling: spikes, events, locality. The two fields share a vocabulary and an ancestry in the 1943 neuron model of McCulloch and Pitts, and little else.