Configurable convolutional array accelerator structure based on Winograd

An accelerator and convolution technology, applied in neural architecture, complex mathematical operations, biological neural network models, etc., can solve problems such as inflexible configuration and reduced applicability

Active Publication Date: 2019-09-27
TIANJIN UNIV
View PDF9 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the convolution data bit width in the existing fixed-point neural n...

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Configurable convolutional array accelerator structure based on Winograd
  • Configurable convolutional array accelerator structure based on Winograd
  • Configurable convolutional array accelerator structure based on Winograd

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0058] A Winograd-based configurable convolution array accelerator structure of the present invention will be described in detail below in conjunction with the embodiments and drawings.

[0059] In the convolution calculation of the neural network, the Winograd conversion formula is

[0060] Out=A T [(GKG T )⊙(B T IB)]A(1)

[0061] Among them, K represents the weight matrix in the time domain, I represents the activation value matrix in the time domain, and A, G, and B represent the result matrix of point multiplication [(GKG T )⊙(B T IB)], the time domain weight matrix K, the conversion matrix corresponding to the time domain activation value matrix I, the conversion matrices A, G, B are specifically as follows:

[0062]

[0063] The output paradigm of the Winograd convolution used in the present invention is F(2*2,3*3), the first parameter 2*2 represents the size of the output feature map, and the second parameter 3*3 represents the size of the convolution kernel . ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a configurable convolutional array accelerator structure based on Winograd. The structure comprises an activation value caching module, a weight caching module, an output caching module, a controller, a weight preprocessing module, an activation value preprocessing module, a weight conversion module, an activation value matrix conversion module, a dot multiplication module, a result matrix conversion module, an accumulation module, a pooling module and an activation module. According to the configurable convolutional array accelerator structure based on the Winograd, the convolutional array accelerator with the configurable bit width is designed according to the operation characteristics of a Winograd convolution algorithm of a fixed normal form, and the requirements of different neural networks and different convolutional layers for the bit widths are flexibly met. In addition, a special multiplier unit with configurable data bit width is also designed, so that the calculation efficiency of the neural network convolution operation is improved, and the calculation power consumption is reduced.

Description

technical field [0001] The invention relates to a configurable convolution array accelerator structure. In particular, it relates to a Winograd-based configurable convolutional array accelerator structure. Background technique [0002] Neural networks perform well in many fields, especially image-related tasks, such as image classification, image semantic segmentation, image retrieval, object detection and other computer vision problems. They have begun to replace most traditional algorithms and are gradually deployed on terminal devices. [0003] However, the amount of calculation of the neural network is very huge, so there are problems such as slow processing speed of the neural network and high power consumption. A neural network mainly includes a training phase and an inference phase. In order to obtain high-precision processing results, weight data needs to be obtained through repeated iterative calculations from massive data during training. In the neural network r...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06N3/063G06N3/04G06F17/16
CPCG06N3/063G06F17/16G06N3/045Y02D10/00
Inventor 魏继增徐文富王宇吉郭炜
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap