Granted Patent B2

US 12,738,952 · App. 18/501,664 · Granted Sep 15, 2026

Inventors: Febin Sunny (Folsom, CA); Saideep Tiku (Folsom, CA); Shashank Bangalore Lakshman (Folsom, CA); Poorna Kale (Folsom, CA)

Assignee: Micron Technology, Inc.

H03M1/282H03M1/26

Abstract

Claims (52)

1 . A method, comprising:

receiving a set of first parameters as an input for an operation of multiplication and accumulation;

analyzing the set of first parameters to identify an encoding parameter;

generating, according to the encoding parameter and the set of first parameters, analog inputs to an analog section of the operation of multiplication and accumulation;

generating, using the analog section responsive to the analog inputs, analog outputs; and

determining, according to the encoding parameter and the analog outputs, a set of second parameters as an output responsive to the input for the operation of multiplication and accumulation;

wherein the analog section has a characteristic of consuming more energy for processing input parameters having smaller magnitudes.

2 . The method of claim 1 , wherein the encoding parameter is identified to increase magnitudes of the analog inputs.

3 . The method of claim 2 , wherein the analog section includes microring resonators configured as computing elements for the operation of multiplication and accumulation in optical domain.

4 . The method of claim 3 , wherein the generating of the analog inputs includes:

generating, using an encoder controlled by the encoding parameter, a set of third parameters from the set of first parameters; and

converting, using digital to analog converters, the set of third parameters to the analog inputs.

5 . The method of claim 4 , wherein the encoder includes bitwise shifters.

6 . The method of claim 4 , wherein the determining of the set of second parameters includes:

converting, using analog to digital converters, the analog inputs to a set of fourth parameters; and

generating, using a decoder controlled by the encoding parameter, the set of second parameters from the set of fourth parameters.

7 . The method of claim 3 , wherein the generating of the analog inputs includes:

converting, using digital to analog converters, the set of first parameters to first signals; and

amplifying the first signals, using amplifiers controlled by the encoding parameter, to generate the analog inputs.

8 . The method of claim 7 , wherein the determining of the set of second parameters includes:

reducing the analog outputs, using attenuators controlled by the encoding parameter, to second signals; and

converting, using analog to digital converters, the second signals to the set of second parameters.

9 . The method of claim 3 , wherein the analyzing of the set of first parameters includes an analysis of leading zeroes of the first parameters.

10 . The method of claim 9 , wherein the analyzing of the set of first parameters includes identifying a statistical parameter of maximum depths of leading zeroes of the first parameters.

11 . The method of claim 10 , wherein the encoding parameter is representative of an average of the maximum depths of leading zeroes of the first parameters.

12 . The method of claim 9 , wherein the analyzing of the set of first parameters includes identifying a selected one of maximum depths of leading zeroes of the first parameters through minimizing an information distance between a second set of parameters to be fed into the microring resonators following a first set of parameters in processing the analog inputs.

13 . A device, comprising:

memory cells configured to store a set of first parameters as an input for an operation of multiplication and accumulation;

digital to analog converters;

an analog section of the operation of multiplication and accumulation;

analog to digital converters; and

a controller configured to:

analyze the set of first parameters to identify an encoding parameter;

generate, using the digital to analog converters and according to the set of first parameters, analog inputs to the analog section of the operation of multiplication and accumulation, wherein the analog section is configured to generate analog outputs responsive to the analog inputs; and

determine, using the analog to digital converters and according to the encoding parameter and the analog outputs, a set of second parameters as an output responsive to the input for the operation of multiplication and accumulation;

wherein the analog section has a characteristic of consuming energy at varying levels for processing input parameters having varying magnitudes.

14 . The device of claim 13 , wherein the encoding parameter is identified to reduce energy expenditure in the analog section in generating the analog outputs.

15 . The device of claim 14 , wherein the analog section includes microring resonators or memristors configured as computing elements for the operation of multiplication and accumulation in optical domain.

16 . The device of claim 15 , wherein the controller is configured to perform a statistical analysis of leading zeroes or trailing zeros of the first parameters to determine the encoding parameter.

17 . The device of claim 15 , wherein the controller is further configured to perform an information analysis to select the encoding parameter to minimize an information distance between a second set of parameters to be fed into the computing elements following a first set of parameters in processing the analog inputs.

18 . A non-transitory computer storage medium storing instructions which, when executed in a computing device, cause the computing device to perform a method, comprising:

receiving a set of first parameters as an input for an operation of multiplication and accumulation;

analyzing the set of first parameters to identify an encoding parameter;

generating, according to the encoding parameter and the set of inputs, a set of third parameters;

providing the set of third parameters to an analog accelerator having an analog section of the operation of multiplication and accumulation;

receiving, from the analog accelerator, a set of fourth parameters generated by the analog accelerator in response to the set of third parameters; and

generating, according to the encoding parameter and the set of fourth parameters, a set of second parameters as an output responsive to the input for the operation of multiplication and accumulation;

wherein the analog section has a characteristic of consuming energy at varying levels for processing input parameters having varying magnitudes.

19 . The non-transitory computer storage medium of claim 18 , wherein the method further comprises:

performing a statistical analysis of leading zeroes or trailing zeros of the first parameters to determine the encoding parameter.

20 . The non-transitory computer storage medium of claim 18 , wherein the method further comprises:

performing an information analysis to select the encoding parameter via minimizing an information distance between a second set of parameters to be fed into computing elements in the analog section following a first set of parameters in processing the set of third parameters.

Assignments (1)

Continuity (2)

References Cited (15)