Benchmarks are a critical tool for evaluating the performance of hardware and software systems. In this survey paper, we discuss the importance of benchmarks in various fields, including computer science and engineering, and hardware security. We define benchmarks and their types, including natural and synthetic benchmarks, and discuss the limitations of existing benchmarks. We also examine different approaches for creating synthetic benchmarks, such as random generation and optimization-based approaches. Furthermore, we highlight the role of logic and layout synthesis tools in characterising bench-marks and discuss the different benchmarks used to evaluate hardware systems' performance, including area, power consumption, timing, manufacturability, signal integrity, and testability. Finally, we suggest strategies for overcoming the limitations of existing benchmarks, such as developing new benchmarks, updating existing ones, increasing benchmark size, and simplifying benchmarks. Overall, this survey paper underscores the importance of benchmarks in research and development and provides insights into their creation, characterization, and application.