This chapter explores the balance between power demand and performance achieved by hardware manufacturers through dynamic voltage and frequency scaling (DVFS) technology. Section 7.1 provides a detailed overview of DVFS and its various governors—onDemand, Conservative, Performance, Powersave, and Userspace—explaining their distinct rules for adjusting processor frequency while presenting mathematical metrics for quantifying performance and power requirements. The chapter in Sect. 7.2 further delves into queuing theory as a key methodology for analyzing computing system performance, outlining essential modeling parameters and discussing single-server, multi-server, and infinite-server systems, along with relevant performance metrics. In Sect. 7.3, the focus shifts to Markovian analysis for gaining insights into the power-performance trade-off, defining Markov processes and illustrating discrete-time and continuous-time Markov chains. The section concludes with methods for conducting effective Markovian analysis. Section 7.4 establishes the appropriateness of the single-server M/M/1 queuing system for analyzing the performance-power trade-off, assuming constant maximum frequency, and confirms its accuracy against real-world metrics. Finally, the chapter in Sect. 7.5 introduces discrete-time Markov chain (DTMC) models for utilization-based governors, particularly onDemand and Conservative, incorporating dynamic scaling factors. It concludes with a detailed discussion on the steady-state probabilities of the DTMC models, enhancing the reader’s understanding of these governors’ dynamics.