Dynamic voltage scaling (DVS) is an effective low-power design technique for embedded real-time systems. In recent years, many DVS algorithms have been proposed for reducing the energy consumption of embedded hard real-time systems. However, the proposed DVS algorithms were not quantitatively evaluated under a unified framework, making it a difficult task to select an appropriate DVS algorithm for a given application/system. In this paper, we compare several key DVS algorithms recently proposed for hard real-time periodic task sets, analyze their energy efficiency, and discuss the performance differences quantitatively. Our evaluation results give quantitative answers to several important DVS questions.