Configurable Computing Array Using 3D LUT Memory for Complex Math

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional configurable gate arrays are limited by fixed computing elements that cannot be reconfigured to perform complex math functions, restricting their application in high-performance applications that require flexible mathematical operations.

Innovation Solution

Incorporating three-dimensional writable memory (3D-W) arrays that can be electrically programmed with look-up tables for math functions, allowing configurable computing elements to realize various math operations and enabling the combination of basic functions to implement complex math functions, along with configurable logic and interconnects.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If fixed computing elements are used in conventional configurable gate arrays, then the device structure is simple and manufacturing is easier, but the adaptability to perform complex math functions is limited

Engineering Contradiction:

Improvemath function capabilityVSAvoidcomputing element structure

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements configurable computing elements that can perform multiple math functions (addition, subtraction, multiplication, division, square root) by loading different look-up tables into 3D-W memory arrays. This allows a single computing element structure to universally handle various mathematical operations, resolving the contradiction between adaptability and complexity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The computing elements transition from fixed functionality to dynamic reconfigurability through electrical programming of the 3D-W memory. The ability to load different look-up tables at runtime enables the system to adapt its mathematical capabilities dynamically, achieving versatility without permanent structural complexity.

Inventive Principle:

Principle #15Dynamics

2Productivity

If fixed computing elements are used, then manufacturing precision requirements are lower, but the computing power and flexibility for high-performance applications are insufficient

Engineering Contradiction:

Improvecomputing powerVSAvoidconfigurable element fabrication

Core Design Contradiction:

ProductivityVSManufacturing precision

Solution Approach 1:

The patent replaces fixed hardware circuitry with electrically programmable 3D-W memory structures. This substitution allows mathematical functions to be defined through data (look-up tables) rather than fixed physical circuit connections, enabling high computing power while using standard semiconductor manufacturing processes.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The computing elements change their functional parameters by loading different look-up tables into the 3D-W memory. This parameter-based configuration approach allows the same physical structure to deliver varying computing capabilities without requiring different manufacturing precision levels for each function.

Inventive Principle:

Principle #35Parameter changes

3Measurement precision

If traditional two-dimensional memory is used for look-up tables, then the device structure is simpler, but the storage density and precision for math functions are reduced

Engineering Contradiction:

Improvemath function precisionVSAvoidmemory structure

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent transitions from two-dimensional memory structures to three-dimensional stacked memory arrays for storing look-up tables. This dimensional enhancement provides exponentially higher storage density, enabling high-precision mathematical functions with sufficient lookup table entries while maintaining compact device footprints.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

4Adaptability or versatility

If more computing elements are added to increase computing power, then the math function capability improves, but the die size and manufacturing costs increase

Engineering Contradiction:

Improvemath function varietyVSAvoiddie size

Core Design Contradiction:

Adaptability or versatilityVSArea of stationary object

Solution Approach 1:

Each computing element is designed as a universal unit capable of performing multiple mathematical operations by loading different look-up tables. This universality allows the system to achieve high math function variety with fewer physical computing elements, reducing die size while maintaining versatility.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The use of three-dimensional stacked memory structures provides high storage density within a small footprint. This enables each computing element to store comprehensive look-up tables for multiple math functions without increasing die area, resolving the contradiction between function variety and device size.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Data Source

PatentUS10456800B2Configurable computing array comprising configurable computing elements

Publication Date: 2019.10.29 SHENZHEN CUNDI TECHNOLOGY CO LTD

AI summary

To implement a complex math function, i.e. a math function with multiple input variables, a configurable computing array comprises at least an array of configurable computing elements. Each configurable computing element comprises at least a memory which stores a look-up table (LUT) for a math function with a single input variable.