**Author details**

Arijit Banerjee

**6. Discussion**

assist and tracking [20].

142 Green Electronics

Energy consumption in billion node IoT networks is expected to increase, as the total no. of IoT devices may reach 50 billion by the year 2020. A portion of these massive numbers of IoT devices will be plugged into the outlets in homes, factories, and outdoor settings. On the other hand, a huge number of IoT devices will be battery-operated or energy-harvested portable systems. The billions of IoT devices plugged into the outlets will draw power from the energy grid resulting in millions of dollars in energy bills and will increase the carbon footprint of this planet. Moreover, with a shorter battery life and replacement time, supporting billions of battery-operated IoT devices will require a massive production of portable batteries increasing the carbon footprint of Earth, too. Reducing the carbon footprint of these IoT devices requires reduction of power consumption, usage of low voltage operation for quadratic energy savings, and harvesting energy from the environment, which will require ULP IoT SoCs to reduce the energy cost and improve the battery life for a greener IoT electronics. However, technology scaling in the latest 7 nm FinFET and beyond will become a hindrance to lower operating voltage of the widely used embedded SRAMs, which shares the same power line with the digital core in ULP IoT SoCs. This chapter reviews some of the state-of-the-art SRAM design techniques, which are promising candidates for reducing power

**Figure 12.** Measured canary VMIN tracking across clock frequencies [1 or 10, 50, 100, and 150] MHz and temperatures (a) 27°C, (b) 85°C, and (c) −20°C, showing VMIN tuning range, and (d) the distribution of overall VMIN reduction using

Address all correspondence to: ab9ca@virginia.edu

University of Virginia, Charlottesville, Virginia, USA
