Hardware architecture for simulating a neural network of neurons
AssigneeInternational Business Machines Corporation
InventorsRodrigo Alvarez-Icaza Rivera, John V. Arthur, Andrew S. Cassidy, Pallab Datta, Paul A. Merolla, Dharmendra S. Modha
Life of this patent
Estimating the term…
Abstract
Embodiments of the invention relate to a neural network system for simulating neurons of a neural model. One embodiment comprises a memory device that maintains neuronal states for multiple neurons, a lookup table that maintains state transition information for multiple neuronal states, and a controller unit that manages the memory device. The controller unit updates a neuronal state for each neuron based on incoming spike events targeting said neuron and state transition information corresponding to said neuronal state.
Claims27
A neural network system, comprising:
- a memory device that maintains neuronal states for multiple neurons;
- a lookup table that maintains state transition information for multiple neuronal states;
- and
- a controller unit that manages the memory device, wherein the controller unit updates a neuronal state for each neuron based on incoming spike events targeting said neuron and state transition information corresponding to said neuronal state.
An apparatus, comprising:
- a memory device including a lookup table that maintains information about multiple system states of a model system, wherein the lookup table is configured to provide information relating to a change in a system state when addressed by a current system state of the model system;
- and
- a controller unit configured to update system states for multiple instances of the model system based on the information maintained in lookup table and incoming data received by the controller unit.
A method, comprising:
- maintaining neuronal states for multiple neurons;
- maintaining state transition information for multiple neuronal states in a lookup table;
- and
- updating a neuronal state for each neuron based on incoming spike events targeting said neuron and state transition information corresponding to said neuronal state.
A method, comprising:
- maintaining system states for multiple instances of a model system;
- maintaining system state change information for said system states in a lookup table, wherein the system state change information comprises information relating to changes in said system states;
- and
- for each instance of the model system, updating the system state for said instance based on incoming data and the system state change information maintained.
Dependencies are inferred from the claim text.