Hybrid dynamic thermal management method with model predictive control

Author

Jian Ma ; Hai Wang ; Tan, Sheldon X.-D ; Chi Zhang ; He Tang

Author_Institution

Sch. of Microelectron. & Solid-State Electron., Univ. of Electron. Sci. & Technol. of China, Chengdu, China

fYear

2014

fDate

17-20 Nov. 2014

Firstpage

743

Lastpage

746

Abstract

Dynamic thermal management methods are important to control the temperature of microprocessor at runtime and improve the reliability performance of the chip. In this paper, a model predictive control based dynamic thermal management method is proposed with hybrid thermal regulation techniques combining task migration and dynamic voltage frequency scaling (DVFS) methods. The new method is able to track the specified safe temperature ceiling, reduce processor performance degradation, and lower temperature variation among different cores.

Keywords

microprocessor chips; multiprocessing systems; power aware computing; predictive control; temperature control; thermal management (packaging); DVFS methods; chip reliability performance; dynamic voltage frequency scaling; hybrid dynamic thermal management method; hybrid thermal regulation techniques; microprocessor temperature control; model predictive control; safe temperature ceiling; task migration; Microprocessors; Power demand; Power distribution; Predictive control; Reliability; Temperature distribution; Thermal management;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems (APCCAS), 2014 IEEE Asia Pacific Conference on

Conference_Location

Ishigaki

Type

conf

DOI

10.1109/APCCAS.2014.7032888

Filename

7032888