Scholarlink

In this chapter, we elaborate on the state-of-the-art run-time mitigation approaches for tackling performance variability. In principle, these approaches aim to run workloads with just enough processor speed and hardware resources that the tasks could finish just in time. In this way, the minimum power and energy consumption can be reached to avoid thermal problems, reduce energy costs, and prolong the duration of battery-charged devices. This chapter is structured as follows: First, we describe the general structure of mitigation approaches for performance variability in Sect. 3.1. Then, in Sect. 3.2 we discuss several typical system knobs used in performance variability mitigation. Finally, Sect. 3.3 discusses the heart of the mitigation approaches, the adaptive control algorithm for scheduling with and without timing guarantees.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Mitigation Approaches for Performance Variability: State-of-the-Art

  • 章节

  • 发表日期：

    2025年

查看全文
  • Ji-Yung Lin,
  • Michalis Noltsis,
  • Dimitrios Soudris,
  • Francky Catthoor

摘要

In this chapter, we elaborate on the state-of-the-art run-time mitigation approaches for tackling performance variability. In principle, these approaches aim to run workloads with just enough processor speed and hardware resources that the tasks could finish just in time. In this way, the minimum power and energy consumption can be reached to avoid thermal problems, reduce energy costs, and prolong the duration of battery-charged devices. This chapter is structured as follows: First, we describe the general structure of mitigation approaches for performance variability in Sect. 3.1. Then, in Sect. 3.2 we discuss several typical system knobs used in performance variability mitigation. Finally, Sect. 3.3 discusses the heart of the mitigation approaches, the adaptive control algorithm for scheduling with and without timing guarantees.