Assessing Energy Efficiency of Actual MCUs
The Internet of Things will oversee the deployment of countless intelligent nodules to evaluate, monitor, report and control different aspects of our world. The issue is, these nodes must be in an ‘always on’ condition.
At the same time, there should be an ‘install and forget’ solution. These nodes have to constantly monitor and respond, to all relevant events immediately even as they are under energy constraints.
For example, if a node is working on a form-factor battery, it will have to make do with the same battery for years. Even if the node is able to harvest energy, it must be capable of operating even when little ambient energy is available for harvesting. Basically, embedded engineers are going to have their hands full.
Moreover, IoT nodes contain a number of semiconductor components like sensors that collect data from the environment (humidity, temperature and more). Some of them are able to communicate with the cloud or other nodes, like smart phones. All this is going to need a lot of energy.
Embedded developers can save significant amounts of energy by duty-cycling the components and ensuring they are on only when needed. Even then, some components cannot be turned fully off, like the micro controller. Here, the embedded designer must create an architecture that enables the MCU to work on low energy even as the MCU keeps full control of the operation.