Optical Computing Circuits Using Time-Multiplexed Light for Low-Power AI

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing electronic circuits face challenges with high power consumption in computing-intensive applications like artificial intelligence and machine learning, which are constrained by physical limits and inefficient data computation frameworks.

Innovation Solution

Implementing optical computing using optical/photonic devices to perform multiply-accumulate operations, replacing electronic MAC units with optical beams and spatial light modulators, and utilizing time-multiplexing to reduce energy consumption and hardware requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Power

If electronic circuits are used for AI computing, then processing capability is achieved, but power consumption becomes excessively high

Engineering Contradiction:

Improveprocessing capabilityVSAvoidpower consumption

Core Design Contradiction:

PowerVSUse of energy by moving object

Solution Approach 1:

The patent replaces electronic computing systems with optical computing systems. Specifically, it uses optical beams to represent data and spatial light modulators to perform computational operations, substituting the electronic field with the optical field to achieve low-power AI acceleration while maintaining processing capability

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the fundamental operating parameter from electrical signals to optical signals. By using light intensity, phase, and polarization states to encode and process information, the system achieves computation with significantly reduced energy consumption compared to traditional electronic circuits

Inventive Principle:

Principle #35Parameter changes

2Use of energy by moving object

If optical computing is implemented, then energy consumption is reduced, but device complexity increases

Engineering Contradiction:

Improveenergy consumptionVSAvoiddevice complexity

Core Design Contradiction:

Use of energy by moving objectVSDevice complexity

Solution Approach 1:

The patent employs spatial light modulators that can be programmed to perform different computational functions. These devices serve multiple purposes: they modulate optical beams for multiplication operations, perform accumulation through interference patterns, and can be reconfigured for different neural network layers, reducing the need for dedicated hardware for each function

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent uses optical copying and interference principles where information is replicated across multiple optical paths and then combined. The optical field naturally copies signals through diffraction and interference patterns, enabling parallel computation without requiring complex electronic routing and switching infrastructure

Inventive Principle:

Principle #26Copying

3Productivity

If traditional electronic MAC units are used, then computation is performed, but hardware requirements and area occupation are excessive

Engineering Contradiction:

Improvecomputation performanceVSAvoidhardware area

Core Design Contradiction:

ProductivityVSArea of stationary object

Solution Approach 1:

The patent transitions from two-dimensional electronic circuit layouts to three-dimensional optical field manipulation. By using the spatial dimensions of light propagation and the temporal dimension of optical pulses, the system performs MAC operations in parallel across multiple spatial modes and time slots, achieving high computation performance with minimal physical footprint

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent segments the computational task into discrete optical pulses that can be processed independently. Each pulse carries specific data information and interacts with spatial light modulators in a controlled sequence, enabling efficient multiplication and accumulation operations without requiring large-scale integrated circuit interconnections

Inventive Principle:

Principle #1Segmentation

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Optical computing significantly reduces energy consumption and enables efficient processing of large neural networks with minimal time delay, compatible with current manufacturing processes and compatible with existing photonic device frameworks.

Implementation Method 1

a first modulator pixel...configured to be modulated, in correspondence with the K first input optical pulses, according to K first weights

Methodology Applied

Scientific EffectOptical modulation: Electro-Optic Effects

Implementation Method 2

a first detector pixel...configured to generate charges in response to the plurality of modulated optical pulses

Methodology Applied

Scientific EffectPhotoelectric effect: Photoelectric Effect

Data Source

PatentUS20250247155A1Optical computing device for artificial intelligence accelerators and method of operating the same

Publication Date: 2025.07.31 TAIWAN SEMICONDUCTOR MANUFACTURING CO LTD

AI summary

An optical circuit includes: a source pixel configured to generate a plurality of input optical pulses with a time interval; a modulator pixel optically coupled to the source pixel and configured to modulate the plurality of input optical pulses to generate a plurality of modulated optical pulses; a detector pixel optically coupled to the modulator pixel and configured to generate charges in response to the plurality of modulated optical pulses; and a controller configured to electrically control an intensity of each of the plurality of input optical pulses and modulation levels of the modulator pixel to perform a multiplication-accumulation operation.