ULPBench takes a two pronged approach to measuring the energy efficiency of an MCU. Firstly, it performs a range of tests to measure the MCU's efficiency using a common set of workloads that allow low power modes to come into play. Secondly, it uses EEMBC's EnergyMonitor to determine how much power the MCU consumed whilst performing the test.
STMicroelectronics kicked things off at embedded world, claiming a score of 123 on the ULPmark scale for its STM32L476 and 486 parts. Texas Instruments was next up, with a score of 167.4 for its recently released MSP432P401R. Now it's Atmel's turn, claiming 185.8 for its SAML21J18A.
When you look at benchmark reports, however, it's clear that apples haven't been compared with apples. For a start, each device uses a different Cortex core: ST has an M4, TI has and M4F and Atmel has the M0+. Neither do they have the same clock speeds. In the tests, the MCUs ran at 24, 24 and 12MHz respectively. You might expect an M0+ based device to consume less power than one with an M4F at its heart.
Nevertheless, it's good to see attention being paid to energy efficiency. Many of the devices into which these MCUs will be designed are likely to be powered by batteries, so anything which makes them last longer is good news for engineers and consumers alike.
