Processor and processing method thereof, chip, chip packaging structure and electronic device

A processor and instruction processing technology, applied in the field of computer architecture and neural network, can solve the problems of high energy consumption, inability to guarantee the real-time processing of video data, etc., and achieve low power consumption, simple and fast video processing, and low memory Effect

Active Publication Date: 2019-01-01
SHANGHAI CAMBRICON INFORMATION TECH CO LTD
View PDF9 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, neural networks face very serious challenges in video applications, because the amount of data contained in a video far exceeds the amount of data contained in a picture, so processing a video will consume a lot of time and consume huge energy consumption And cannot guarantee the real-time performance of processing video data

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
  • Processor and processing method thereof, chip, chip packaging structure and electronic device
  • Processor and processing method thereof, chip, chip packaging structure and electronic device
  • Processor and processing method thereof, chip, chip packaging structure and electronic device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0128] This embodiment proposes a processor, which includes a video decoding unit, a storage unit, an instruction control unit, an accurate calculation unit, and an approximate calculation unit.

[0129] Wherein, the storage unit pre-stores neural network parameters such as processing instructions and weights.

[0130] like Figure 4 As shown, the processing method of the processor includes the following steps:

[0131] S1. The video decoding unit decompresses the compressed video, extracts the frame information of each frame of the picture in the video, and divides the frame information into complete frame information and reference frame information;

[0132] S2. The storage unit stores the complete frame information and reference frame information divided by the video decoding unit;

[0133] S3. The instruction control unit extracts the processing instruction of the storage unit; and decodes the processing instruction to control the precise calculation unit and the approxi...

Embodiment 2

[0146] This embodiment proposes a processor, which includes a video decoding unit, a storage unit, an instruction control unit, an accurate calculation unit, and an approximate calculation unit, wherein:

[0147] The processor is applied to the acceleration device to accelerate the detection of video objects, specifically:

[0148] The video decoding unit extracts and classifies the frame information of each frame of the video while decompressing the video. The frame information of the video can be divided into three types. Contains the encoded frame information of the difference part - P frame, and refers to the encoded frame information of the previous and subsequent frames - B frame.

[0149] The video decoding unit gives each frame information of the video a mark bit to mark the type of the frame information. For the P frame, the video decoding unit will also mark the forward reference frame of the frame, and will also extract the object passing through the P frame. Displ...

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 processor. The processor includes an instruction control unit and a calculating unit; the instruction control unit is used for extracting a processing instruction to controlthe calculating unit; and the calculating unit is used for executing neural network operation based on input frame information and neural network parameters. The processor realizes efficient functional reconstruction on neural network processors, and can give full play to performance in application environments with low memory and strong real-time performance.

Description

technical field [0001] The disclosure belongs to the field of computer architecture and neural network, and more specifically relates to a processor and its processing method, a chip, a chip packaging structure, a card board and an electronic device. Background technique [0002] Neural network (neural network) has not only made outstanding achievements in the field of image applications in recent years, but also has outstanding performance in the field of video applications. However, neural networks face very serious challenges in video applications, because the amount of data contained in a video far exceeds the amount of data contained in a picture, so processing a video will consume a lot of time and consume huge energy consumption And the real-time performance of processing video data cannot be guaranteed. Therefore, how to reduce the time for neural networks to process video has become an urgent problem to be solved. Contents of the invention [0003] Based on the ...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06N3/06
CPCG06N3/06G06V20/46G06V10/454G06V10/82G06N3/048G06F9/3001G06F9/30094G06F9/30145G06N3/02
Inventor 不公告发明人
Owner SHANGHAI CAMBRICON INFORMATION TECH CO LTD
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