基于多核微处理器温度感知的线程调度算法

Temperature-aware thread scheduling algorithm for multi-core processors

由于多核微处理消耗更多的能量，导致其热点数目增加，温度分布不平衡加剧，因而对性能产生更大的负面影响。为了解决这个问题，提出一种基于多核微处理器温度感知的线程调度算法来减少热紧急事件、提高性能，并在一个Intel的多核微处理器平台上实现了该算法。实验结果表明，在各种负载组合下，该算法可以减少9．6％-78．5％的动态热管理次数。与Linux标准调度算法相比，吞吐率平均可以提高5．2％，最大可提高9．7％。

Muhi-core microprocessor consumes more power, leading to an increase in the number of its hot spots and uneven distribution of temperature, which creates a greater negative impact on performance. Concerning this problem, a temperature-aware thread scheduling algorithm was proposed to reduce the thermal emergencies and improve the throughput, which had been implemented, in the Linux kernel on Intel＇s Quad-Core system. The experimental results show that this schedule algorithm can reduce 9.6% - 78.5% of thermal emergencies in various combinations of workloads, and has an average of 5.2% and up to 9.7% throughput higher than the Linux standard scheduler.