The hardware cryptography blocks add to the number of functions that the chip can perform without waking the ARM cortex M3/M4 cores from sleep; these include making measurements and carrying out communications operations, and are a cornerstone of the low-power claims made for the chips.
The two EFM32 Gecko microcontroller (MCU) families, Jade Gecko and Pearl Gecko, host a hardware cryptography engine, flexible low-energy modes, an on-chip dc-dc converter and scalable memory options; you program them in the company’s Simplicity Studio toolset where there are comprehensive features for energy management. EFM32 Jade and Pearl Gecko MCUs target an array of energy-sensitive and battery-powered devices including wearable health and fitness trackers, smart door locks, point-of-sale devices, security sensors and other IoT node applications.
Silcon labs says that Jade and Pearl Gecko MCUs address the growing need to equip IoT-connected devices with the advanced security technologies to thwart hackers. The hardware cryptography engine promises fast, energy-efficient, autonomous encryption and decryption for Internet security protocols such as TLS/SSL with minimal CPU intervention—and without sacrificing battery life. The on-chip crypto-accelerator supports advanced algorithms such as AES with 128- or 256-bit keys, elliptical curve cryptography (ECC), SHA-1 and SHA-224/256. Hardware cryptography enables developers to meet evolving IoT security requirements more efficiently than with conventional software-only techniques.
Based respectively on ARM Cortex-M3 and M4 cores, Jade and Pearl Gecko have what Silicon labs terms an enhanced peripheral reflex system (PRS) – this is a communications structure that allows peripheral functions to operate and exchange data without powering-up the core. It is controlled by a state-machine-like system that your code configures before the main core of the device enters deep-sleep mode: it is capable of simple logical decision making to extend the range of what can be done before restarting the ARM core. Low active-mode current (63 µA/MHz) enables computationally intensive tasks to execute faster. Low sleep-mode current (1.4 µA down to 30 nA) and ultra-fast wake-up/sleep transitions further minimise energy consumption. When using Silicon Labs’ own tool set, and its development boards which include current-measurement in the power line, you can correlate code events to actual power consumed, to optimise routines for minimum energy. The software integrates current measurements to yield a realistic indication of energy taken from a battery supply.
Jade and Pearl Gecko MCUs also integrate a high-efficiency dc-dc buck converter. Offering a total current capacity of 200 mA, the on-chip converter can provide a power rail for other system components in addition to powering the MCU. This power management innovation reduces bill-of-materials (BOM) cost and board area by eliminating the need for an external dc-dc converter. The (switching) buck converter has been designed, Silicon labs assures, to operate without generating noise that could disturb analogue functions.
Features include; 32-bit ARM Cortex-M3 and M4 cores scaling up to 40 MHz; on-chip hardware encryption/decryption accelerator supporting random number generator, AES, ECC and SHA technologies; energy management system with a wide range of low-energy modes; 12-channel peripheral reflex system (PRS) supporting autonomous interaction between peripherals down to deep sleep mode, as well as configurable logic functions; 2-bit, 1 Msample/sec ADC capable of sampling and performing autonomous comparisons down to stop mode; memory options up to 256 kB flash with 32 kB RAM; and an independent I/O rail supporting 1.8V to 5V components.
Engineering samples of EFM32JG Jade Gecko and EFM32PG Pearl Gecko MCUs are available in 5 x 5 mm QFN32 and 7 x 7 mm QFN48 packages; production quantities are planned for Q2 2016. Jade Gecko pricing begins at $1.24 Pearl Gecko at $1.65 (both 10,000). The SLSTK3401A EFM32PG Pearl Gecko Starter Kit is priced at $29.99.
Silicon Labs; www.silabs.com/EFM32