Brain-like computing system and computing method based on neuron instruction coding

Abstract

The invention discloses a brain-like computing system and computing method based on neuron instruction coding, including an impulse neural network, wherein the impulse neural network includes a plurality of neuron nodes; and also includes: a neuron instruction encoding unit, which is used for coding The method specifies at least one of the parameters, parameter arrangement and parameter use methods required for the execution of the neuron instruction; the neuron node participating in the calculation executes the calculation according to the neuron instruction formed by the neuron instruction coding unit and combined with its own parameter values. The brain-like computing system and computing method realize the computing control of different neurons by coding neuron instructions, have both the specificity and versatility of computing functions of each neuron cluster, and can realize extended computing of different neurons.

Description

technical field [0001] The invention belongs to the technical field of impulse neural networks, and in particular relates to a brain-like computing system and computing method based on neuron instruction coding. Background technique [0002] In recent years, the effects of "memory wall" and "power wall" have become increasingly serious, and the von Neumann architecture followed by traditional computers is facing huge challenges. The semiconductor industry urgently needs to find new architectures and approaches to meet the electronics industry's demands for ever-increasing computing performance and extremely low power consumption. With the development of brain science, people gradually understand that the human brain is an extremely energy-efficient computer, and brain-like computing emerges as the times require. For example, a brain-like computing system disclosed in patent document CN113222134A. Combining the computing units into one fundamentally removes the "memory wall"...

AI Technical Summary

Problems solved by technology

Although this type of brain computer can support the calculation of hundreds of millions of neurons, it cannot realize the specificity of the calculation function of each neuron cluster.

Images