Systems and methods for device thermal management using dynamic ranking of device cooling needs

Patent metadata

Field	Value
Publication number	US-12164278-B2
Application number	US-202017092812-A
Country	US
Kind code	B2
Filing date	Nov 9, 2020
Priority date	Nov 9, 2020
Publication date	Dec 10, 2024
Grant date	Dec 10, 2024

Abstract

Techniques described herein relate to methods and systems for thermal management of a thermal environment. The method may include using thermal data items from computing devices and time series analysis to predict future thermal values for the thermal data items; performing a clustering analysis using the predicted future thermal values to assign cluster labels to the computing devices; using the cluster labels and the predicted future thermal values to assign predicted thermal status labels to the computing devices; assigning a confidence value to the predicted thermal status labels and ranking the computing devices based on the confidence values; performing an analysis to determine a thermal data item contributing to the assigned thermal status; and sending the results to a thermal environment administrator.

First claim

What is claimed is:

1. A method for thermal management of a thermal environment, the method comprising:
obtaining a plurality of thermal data items associated with a plurality of computing devices in the thermal environment;
writing a plurality of entries in a time series database, the plurality of entries comprising the plurality of thermal data items;
performing a time series analysis to predict a plurality of predicted future thermal values based on the plurality of entries in the time series database;
performing a clustering analysis using the plurality of predicted future thermal values to apply a first cluster label to a first portion of the plurality of computing devices and a second cluster label to a second portion of the plurality of computing devices, wherein the clustering analysis comprises a hierarchical density-based spatial clustering of applications with noise technique to organize the plurality of computing devices into a first cluster and a second cluster;
performing a thermal prediction analysis using the first cluster label, the second cluster label, and the plurality of predicted future thermal values to assign a thermal status label to each of the plurality of computing devices, wherein the thermal status label is one selected from a group consisting of high and low;
performing a confidence analysis to determine a confidence value for the thermal status label assigned to each of the plurality of computing devices;
ranking the plurality of computing devices based on the confidence analysis to obtain a ranked thermal status list;
performing a cause analysis to determine at least one thermal data item contributing to each thermal status label assigned to the plurality of computing devices; and
providing, via a graphical user interface, the ranked thermal status list and results of the cause analysis to an administrator of the thermal environment, wherein, in response to receiving the ranked thermal status list and the results, the administrator initiates a remedial action.

2. The method of claim 1 , wherein, before ranking the plurality of computing devices based on the confidence analysis to obtain the ranked thermal status list, at least one thermal status label for a computing device of the plurality of computing devices is discarded for having the confidence value below a confidence value threshold.

3. The method of claim 1 , wherein the plurality of entries in the time series database comprises a chronological series of a plurality of portions of the plurality of thermal data items for each of the plurality of computing devices.

4. The method of claim 1 , wherein the thermal prediction analysis comprises using a Mondrian forest classifier.

5. The method of claim 1 , wherein the confidence analysis comprises using a transductive confidence machine.

6. The method of claim 5 , wherein the cause analysis comprises using an exchangeability test and the plurality of thermal data items.

7. The method of claim 6 , wherein the exchangeability test comprises an analysis using martingales.

8. The method of claim 6 , wherein, when the cause analysis yields a cause result over a threshold, the at least one thermal data item is selected.

9. The method of claim 1 , wherein the results comprise a computing device identifier, the thermal status label predicted for a computing device having the computing device identifier, a rank of the computing device among the plurality of computing devices, and the at least one thermal data item.

10. A non-transitory computer readable medium comprising computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for thermal management of a thermal environment, the method comprising:
obtaining a plurality of thermal data items associated with a plurality of computing devices in the thermal environment;
writing a plurality of entries in a time series database, the plurality of entries comprising the plurality of thermal data items;
performing a time series analysis to predict a plurality of predicted future thermal values based on the plurality of entries in the time series database;
performing a clustering analysis using the plurality of predicted future thermal values to apply a first cluster label to a first portion of the plurality of computing devices and a second cluster label to a second portion of the plurality of computing devices, wherein the clustering analysis comprises a hierarchical density-based spatial clustering of applications with noise technique to organize the plurality of computing devices into a first cluster and a second cluster;
performing a thermal prediction analysis using the first cluster label, the second cluster label, and the plurality of predicted future thermal values to assign a thermal status label to each of the plurality of computing devices, wherein the thermal status label is one selected from a group consisting of high and low;
performing a confidence analysis to determine a confidence value for the thermal status label assigned to each of the plurality of computing devices;
ranking the plurality of computing devices based on the confidence analysis to obtain a ranked thermal status list;
performing a cause analysis to determine at least one thermal data item contributing to each thermal status label assigned to the plurality of computing devices; and
providing, via a graphical user interface, the ranked thermal status list and results of the cause analysis to an administrator of the thermal environment, wherein, in response to receiving the ranked thermal status list and the results, the administrator initiates a remedial action.

11. The non-transitory computer readable medium of claim 10 , wherein, before ranking the plurality of computing devices based on the confidence analysis to obtain the ranked thermal status list, at least one thermal status label for a computing device of the plurality of computing devices is discarded for having the confidence value below a confidence value threshold.

12. The non-transitory computer readable medium of claim 10 , wherein the plurality of entries in the time series database comprises a chronological series of a plurality of portions of the plurality of thermal data items for each of the plurality of computing devices.

13. The non-transitory computer readable medium of claim 10 , wherein the thermal prediction analysis comprises using a Mondrian forest classifier.

14. The non-transitory computer readable medium of claim 10 , wherein the confidence analysis comprises using a transductive confidence machine.

15. The non-transitory computer readable medium of claim 14 , wherein the cause analysis comprises using an exchangeability test and the plurality of thermal data items, and wherein the exchangeability test comprises an analysis using martingales.

16. The non-transitory computer readable medium of claim 10 , wherein, when the cause analysis yields a cause result over a threshold, the at least one thermal data item is selected.

17. The non-transitory computer readable medium of claim 10 , wherein the results comprise a computing device identifier, the thermal status label predicted for a computing device having the computing device identifier, a rank of the computing device among the plurality of computing devices, and the at least one thermal data item.

18. A system for thermal management of a thermal environment, the system comprising:
a thermal data collector, comprising circuitry, and configured to:
obtain a plurality of thermal data items associated wi

Assignees

• Emc Ip Holding Co Llc

Inventors

• Vishwakarma Rahul Deo
• Gaikwad Hemant
• Singh Ashish

Classifications

• G06N20/20

  Ensemble learning · CPC title

• G05B19/406

  characterised by monitoring or safety (G05B19/19 takes precedence) · CPC title

• G05B2219/49216

  Control of temperature of processor · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06F11/3065

  Monitoring arrangements determined by the means or processing involved in reporting the monitored data (error or fault reporting or logging G06F11/0766) · CPC title