In addition to enabling developers to manage processor and peripheral states, the mbed power management APIs are designed with real-world, low-energy application scenarios in mind. A new feature exposed by the APIs on Silicon Labs’ EFM32 Gecko MCUs automatically determines and enables the optimal sleep mode based on the MCU peripherals in use, which can dramatically reduce system-level energy consumption. Low-energy optimisation is achieved by enabling I/O operations to be executed in the background and by allowing those operations to continue even while the MCU core is in sleep mode or during other processing tasks.
The automatic selection of the optimal sleep mode, combined with low-energy, autonomous MCU peripherals, enables developers to significantly reduce the energy consumption of their IoT applications with minimal effort. For example, energy profiles of an application updating a clock display every second on a memory LCD, a common use case for IoT devices, have shown a current consumption reduction from 1.03mA to 0.100mA.
“The power management APIs for ARM mbed make it possible for developers to create applications that take advantage of the low-power features of ARM Cortex-M based MCUs,” said Zach Shelby, Vice President, IoT Business Marketing, ARM. “This is an important step toward enabling full energy-awareness in IoT devices, and it is one of the key building blocks for mbed OS that is due for public release later this year.”
“As pioneers in low-energy processing solutions for the IoT, Silicon Labs and ARM have made enormous progress in defining and delivering the power management APIs for mbed,” commented Daniel Cooley, Vice President and General Manager, MCU and Wireless Products, Silicon Labs. “We’re excited to help deliver the industry’s first low-power mbed platform, which will play a key role in accelerating the deployment of countless battery-powered IoT applications.”
Silicon Labs plans to provide mbed-enabled EFM32 Gecko starter kits in April 2015. The initial platforms supporting mbed will include the Wonder Gecko, Leopard Gecko, Giant Gecko and Zero Gecko starter kits. Developers with existing EFM32 kits will be able to mbed-enable their hardware through a simple software update.