En cours de chargement...
Following the end of Moore's Law and the emergence of new applicational needs such as those arising from the Internet of Things (IoT) or from Artificial Intelligence (AI), neuro-inspired, or neuromorphic, information processing is attracting increasing attention from the scientific community. The principle of neuro-inspired information processing is to emulate, in a simplified fashion, that most formidable of information processing machines, the brain, using artificial neurons and synapses organized into a network.
These networks can be pieces of software, and can thus be installed in the form of a computer program, but they can also be pieces of hardware and constructed using nanoelectronic circuits. This "hardware" route enables a very low energy usage, the possibility of faithfully reproducing the shape and the dynamics of the action potential in living neurons (biomimetic approach), or even the possibility of being up to a thousand times faster (high-frequency approach).
This is a very promising route, and the proposal by major nanoelectronics manufacturers of having circuits incorporating several million artificial neurons and synapses is not a pipedream.