Softmax Hardware Mapping for Numerically Stable Neural Accelerators

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Implementing a softmax layer in neural network accelerators (NNAs) is challenging due to its complexity and numerical instability issues, especially when dealing with large input values, which can lead to overflow or underflow, and existing NNAs are not designed to efficiently perform this operation.

Innovation Solution

The softmax layer is implemented using a series of elementary neural network operations, including maximum, subtraction, exponential, and division, which are mapped to existing hardware modules in NNAs, such as convolution engines and activation units, to perform the necessary calculations efficiently.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If softmax layer is implemented using traditional dedicated hardware, then computational accuracy is improved, but device complexity and hardware cost increase

Engineering Contradiction:
Improvecomputational accuracyVSAvoidhardware complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements softmax layer using existing universal hardware components (convolution engines, activation units, pooling units) that can perform multiple functions. Instead of dedicated softmax hardware, the system reuses standard NNAs components to achieve the same computational accuracy, thereby reducing hardware complexity and cost while maintaining functionality.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Adaptability or versatility

If softmax layer is implemented using existing NNA hardware, then hardware utilization is improved, but computational efficiency deteriorates

Engineering Contradiction:
Improvehardware utilizationVSAvoidcomputational efficiency
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent segments the softmax layer computation into distinct operational steps (maximum value calculation, subtraction, exponential, division) that can be mapped to different existing hardware modules. This segmentation allows each step to be optimized and executed by appropriate existing components, improving overall computational efficiency while maximizing hardware utilization.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system dynamically maps different parts of the softmax computation to different hardware modules based on the specific computational requirements. The controller flexibly assigns operations to convolution engines, activation units, or pooling units as needed, enabling adaptive hardware utilization that maintains high computational efficiency across different scenarios.

Inventive Principle:
Principle #15Dynamics

3Device complexity

If softmax layer is implemented without dedicated hardware, then device complexity is reduced, but computational performance deteriorates

Engineering Contradiction:
Improvehardware complexityVSAvoidcomputational performance
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent ensures continuous and efficient computation by mapping softmax operations to existing hardware that can process data continuously. The segmented approach allows overlapping operations across different hardware modules, maintaining productive action flow without the need for dedicated hardware, thus achieving both low complexity and high performance.

Inventive Principle:
Principle #20Continuity of useful action

4Adaptability or versatility

If softmax layer is implemented using elementary operations mapping, then hardware utilization is improved, but implementation complexity increases

Engineering Contradiction:
Improvehardware utilizationVSAvoidimplementation complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system uses a controller that automatically handles the mapping of softmax operations to appropriate hardware modules based on the computational graph. This self-service approach abstracts the implementation complexity from the user, allowing high hardware utilization through automated resource allocation while keeping the user interface simple and the overall system manageable.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP4050522B1Implementation of softmax and exponential in hardware
Publication Date: 2025.12.10 IMAGINATION TECH LTD
  • EP4050522B1 patent drawingFigure 1A
  • EP4050522B1 patent drawingFigure 1B
  • EP4050522B1 patent drawingFigure 2

AI summary

Methods are disclosed for implementing an exponential operation, and a softmax neural network layer, in neural network accelerator hardware. Also disclosed are a data processing system for implementing the exponential operation and a data processing system for implementing the softmax layer. The exponential operation or softmax layer is mapped to a plurality of elementary neural network operations, and the neural network accelerator hardware evaluates these operations, to produce the result of the operation or layer respectively.