부산대학교 도서관

Integrated circuits are the backbone of the electronic industry. In which CMOS played an important role for many years. CMOS technology resulted in efficient circuit design but when it comes to area and power, consumption was high in comparison to beyond CMOS devices such as FinFET. After much research and simulations done on FinFET devices, it is proven to be quite promising in terms of area and energy consumption. With rising demand to alter the size of chips to acquire a greater number of transistors in small areas, FinFET has already replaced CMOS for digital circuit design. Similarly, approximate computing is growing parallelly, approximate compressors are important elements in this field which follow the principle of obtaining similar outputs instead of being accurate. Employing an approximate compressor instead of an accurate one for the application such as image processing in this application error can’t be noticed by the human eye and leads us to reduce energy and cross-sectional area of the device. Leveraging FinFET devices to perform logic operations in approximate compressors can take approximate computing to the next level. In this research paper, a 4:2 approximate compressor is proposed which is designed using FinFET, where evaluation of the circuit is made using relevant measurements such as output delay, error rate, and energy consumption. To acquire a better output error rate, it is reduced to 37%. On simulating this circuit 71.8% improvement in delay and 13% improvement in energy consumption can be seen by comparing it to previous work. Outputs of the proposed circuit were obtained by leveraging the HSPICE simulation tool.