Dynamic thermal management by greedy scheduling algorithm

2012 ◽  
Vol 19 (1) ◽  
pp. 193-199 ◽  
Author(s):  
Shuang-xi Qu ◽  
Min-xuan Zhang ◽  
Guang-hui Liu ◽  
Tao Liu
2014 ◽  
Vol 556-562 ◽  
pp. 3431-3437 ◽  
Author(s):  
Jian Jun Zhang ◽  
Tian Hong Wang ◽  
Yu Zhuo Wang

Effective task scheduling is crucial for achieving high performance in heterogeneous computing environments. Whiling scheduling Out-Tree task graphs, many previous heterogeneity based heuristic algorithms usually require high scheduling costs and may not deliver good quality schedules with lower costs. Aiming at the characteristics of Out-Tree task graphs and the features of heterogeneous environments and adopting the strategy based on expected costs and task duplications, this paper proposes a greedy scheduling algorithm, which, at each scheduling step, tries to guarantee not to increase the schedule length, schedules the current task onto the used processor which minimizes its execution finish time; meanwhile, takes load balances into account to economize the use of processors. The comparative experimental results show that the proposed algorithm has higher scheduling efficiency and robust performance, which could produce better schedule which has shorter schedule length and less number of used processors.


2010 ◽  
Vol 54 (10) ◽  
pp. 1584-1606 ◽  
Author(s):  
Anatolij Zubow ◽  
Daniel Camps Mur ◽  
Xavier Perez Costa ◽  
Paolo Favaro

2012 ◽  
Vol 56 (15) ◽  
pp. 3511-3530 ◽  
Author(s):  
Anatolij Zubow ◽  
Johannes Marotzke ◽  
Daniel Camps-Mur ◽  
Xavier Perez Costa

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Tingting Du ◽  
Zixin Xiong ◽  
Luis Delgado ◽  
Weizhi Liao ◽  
Joseph Peoples ◽  
...  

AbstractThermal switches have gained intense interest recently for enabling dynamic thermal management of electronic devices and batteries that need to function at dramatically varied ambient or operating conditions. However, current approaches have limitations such as the lack of continuous tunability, low switching ratio, low speed, and not being scalable. Here, a continuously tunable, wide-range, and fast thermal switching approach is proposed and demonstrated using compressible graphene composite foams. Large (~8x) continuous tuning of the thermal resistance is achieved from the uncompressed to the fully compressed state. Environmental chamber experiments show that our variable thermal resistor can precisely stabilize the operating temperature of a heat generating device while the ambient temperature varies continuously by ~10 °C or the heat generation rate varies by a factor of 2.7. This thermal device is promising for dynamic control of operating temperatures in battery thermal management, space conditioning, vehicle thermal comfort, and thermal energy storage.


Sign in / Sign up

Export Citation Format

Share Document