Supercharge Your Innovation With Domain-Expert AI Agents!

A temperature reconstruction method for many-core chips based on correlation and artificial neural network

An artificial neural network and correlation technology, applied in the field of many-core chip temperature reconfiguration, can solve the problem of high cost and achieve the effect of small number, good flexibility and good practical application value

Active Publication Date: 2021-12-17
NANJING UNIV
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But as the number of cores increases, doing so will make the cost too high

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
  • A temperature reconstruction method for many-core chips based on correlation and artificial neural network
  • A temperature reconstruction method for many-core chips based on correlation and artificial neural network
  • A temperature reconstruction method for many-core chips based on correlation and artificial neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0055] In this embodiment, a method for temperature reconstruction of many-core chips based on correlation and artificial neural network is proposed, which is used to mine the correlation coefficient between cores and core temperatures through artificial neural networks. Intrinsic relationship among them, so as to realize the temperature reconstruction of the many-core chip, and further realize the use of a small number of sensors to obtain the temperature of the whole core chip, which is specifically divided into the following steps:

[0056] Step 1, forming a nuclear temperature measurement group by at least one thermal sensor;

[0057] Step 2, judging the correlation between the cores of the many-core chip and determining the distribution position of the thermal sensor according to the correlation result;

[0058] Step 3, constructing an artificial neural network for temperature reconstruction;

[0059] Step 4, using the core temperature measurement group to collect the te...

Embodiment 2

[0063] In a further embodiment based on the first embodiment, a correlation analysis method is proposed to determine the distribution position of the thermal sensors, so as to solve the problem of how to arrange the limited thermal sensors when the number of cores to be measured is large .

[0064] In the spatial layout, because the cores are physically adjacent to each other, or the points in some special positions have similar sensitivities to the routing algorithm, the core temperature changes in many-core chips have a strong correlation. In order to determine the distribution position of the thermal sensor, the correlation coefficient between cores in the many-core chip is calculated first, and then all the cores are included in the "core set φ where the thermal sensor needs to be placed", and the "core-high Correlation Kernel Matrix ",in is the total number of kernels whose correlation with the corresponding kernel in φ is higher than the preset value, according to th...

Embodiment 3

[0069] In a further embodiment based on the first embodiment, an artificial neural network for temperature reconstruction is proposed, which is used to receive the temperature data collected by the thermal sensor, and obtain the temperature without the thermal sensor core according to the received temperature data data.

[0070] Specifically, the artificial neural network in this example includes two hidden layers, an input layer and an output layer. The layers are connected in a fully connected manner. The number of neurons in the hidden layer is equal to the total number of cores in the many-core chip. Adaptation, the output of the hidden layer is output through a linear rectification function, the input data is the temperature value of the core measured by the thermal sensor, and the output data is the temperature of all cores of the many-core chip.

[0071] In order to better improve the accuracy of the artificial neural network, the loss function is used to adjust the wei...

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 present invention proposes a method and system for reconfiguring the temperature of many-core chips based on correlation and artificial neural network. The method determines the distribution position of thermal sensors based on the correlation between the temperature changes between the cores of many-core chips; The artificial neural network can obtain the temperature of all cores in the superior core chip according to the temperature of some cores obtained by the thermal sensor. The invention can support many-core chips to perform temperature reconfiguration under various load conditions; when the many-core chips run multiple different applications, it is not necessary to re-determine the distribution position of the thermal sensor and retrain the artificial neural network. The invention requires a small number of thermal sensors, and the temperature obtained by reconstruction is high in accuracy, and can adapt to various load situations, so it has good practical value and wide application prospect.

Description

technical field [0001] The invention relates to a temperature reconfiguration method of many-core chips based on correlation and artificial neural network, in particular to the technical field of temperature reconfiguration of many-core chips. Background technique [0002] With the development of semiconductor process technology, the size of devices can be made smaller and smaller, and the number of transistors that can be integrated on a single chip is also increasing. However, due to the limitations of power consumption and temperature, the improvement of chip performance often cannot reach expected. In order to further improve the performance of the chip, multiple processor cores or functional modules are often integrated on the current chip. However, as the number of cores increases, especially when the three-dimensional architecture is proposed, the power density of the chip becomes larger and the heat dissipation path becomes longer, which makes it easy to overheat. ...

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 Patents(China)
IPC IPC(8): G06F1/20G06F15/173G06N3/08
CPCG06N3/08G06F15/173G06F1/206
Inventor 傅玉祥郭孟豪李丽程童何书专李伟
Owner NANJING UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com