TIME-DIVISION MULTIPLEXED NEUROSYNAPTIC MODULE WITH IMPLICIT MEMORY ADDRESSING FOR IMPLEMENTING A NEURAL NETWORK

• US 20140180987A1
• Filed: 12/21/2012
• Published: 06/26/2014
• Est. Priority Date: 12/21/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Arthur, John V., Brezzo, Bernard V., Chang, Leland, Friedman, Daniel J., Merolla, Paul A., Modha, Dharmendra S., Montoye, Robert K., Seo, Jae-sun, Tierno, Jose A.
• Granted Patent: US 9,239,984 B2
• Time in Patent Office: Days
• US Class Current: 706/25
• CPC Class Codes: G06N 3/049 Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
• CPC Class Codes: G06N 3/063 using electronic means
• CPC Class Codes: G06N 3/08 Learning methods

Abstract

Embodiments of the invention relate to a time-division multiplexed neurosynaptic module with implicit memory addressing for implementing a neural network. One embodiment comprises maintaining neuron attributes for multiple neurons and maintaining incoming firing events for different time steps. For each time step, incoming firing events for said time step are integrated in a time-division multiplexing manner. Incoming firing events are integrated based on the neuron attributes maintained. For each time step, the neuron attributes maintained are updated in parallel based on the integrated incoming firing events for said time step.

Claims

1. A method, comprising:

maintaining neuron attributes for multiple neurons; and

maintaining incoming firing events for different time steps;

for each time step, integrating incoming firing events for said time step in a time-division multiplexing manner, wherein said incoming firing events are integrated based on the neuron attributes maintained; and

for each time step, updating the neuron attributes maintained for said multiple neurons based on the integrated incoming firing events for said time step.

2. The method of claim 1, further comprising:

maintaining synaptic connectivity information for said multiple neurons, wherein incoming firing events are integrated based on the synaptic connectivity information maintained.

3. The method of claim 2, further comprising:

dividing said time step into multiple time slots.

4. The method of claim 3, further comprising:

for each time slot, integrating incoming firing events targeting a corresponding incoming axon.

5. The method of claim 4, wherein maintaining neuron attributes for multiple neurons comprises:

for each neuron, maintaining a membrane potential variable and a threshold parameter for said neuron.

6. The method of claim 5, further comprising:

maintaining routing information for said multiple neurons.

7. The method of claim 6, further comprising:

for each neuron, generating an outgoing firing event when a membrane potential variable of said neuron exceeds a threshold parameter of said neuron.

8. The method of claim 7, further comprising:

routing an outgoing firing event generated by each neuron based on the routing information maintained for said neuron.

9. A neurosynaptic device, comprising:

a memory device that maintains neuron attributes for multiple neurons;

a scheduler that manages incoming firing events for different time steps; and

a multi-way processor that integrates incoming firing events for each time step in a time-division multiplexing manner, and updates the neuron attributes maintained for said multiple neurons;

wherein said incoming firing events are integrated based on the neuron attributes maintained.

10. The neurosynaptic device of claim 9, wherein:

the processor maintains synaptic connectivity information for said multiple neurons, wherein incoming firing events are integrated based on the synaptic connectivity information maintained.

11. The neurosynaptic device of claim 10, wherein:

each time step is divided into multiple time slots.

12. The neurosynaptic device of claim 11, wherein:

the neuron attributes for each neuron comprises a membrane potential variable and a threshold parameter for said neuron.

13. The neurosynaptic device of claim 12, wherein:

routing information for said multiple neurons is maintained in a routing lookup table.

14. The neurosynaptic device of claim 13, wherein:

for each neuron, the processor generates an outgoing firing event when a membrane potential variable of said neuron exceeds a threshold parameter of said neuron.

15. A computer program product for a neurosynaptic device, the computer program product comprising a computer-readable storage medium having program code embodied therewith, the program code being executable by a computer to:

maintain neuron attributes for multiple neurons;

maintain incoming firing events for different time steps;

for each time step, integrate incoming firing events for said time step in a time-division multiplexing manner, wherein said incoming firing events are integrated based on the neuron attributes maintained; and

for each time step, update the neuron attributes maintained for said multiple neurons based on the integrated incoming firing events for said time step.

16. The program code of claim 15, further executable by the computer to:

maintain synaptic connectivity information for said multiple neurons, wherein incoming firing events are integrated based on the synaptic connectivity information maintained.

17. The program code of claim 16, further executable by the computer to:

divide said time step into multiple time slots.

18. The program code of claim 17, further executable by the computer to:

for each time slot, integrate incoming firing events targeting a corresponding incoming axon.

19. The program code of claim 18, further executable by the computer to:

for each neuron, maintain a membrane potential variable and a threshold parameter for said neuron; and

maintain routing information for said multiple neurons.

20. The program code of claim 19, further executable by the computer to:

for each neuron, generate an outgoing firing event when a membrane potential variable of said neuron exceeds a threshold parameter of said neuron.