Site Search

Silicon Labs EFM32: Why is Processing time important as a performance metric for low-power MCUs?

EFx32

If you want to reduce the power consumption of the MCU, it is common to design the software so that it operates intermittently. Intermittent operation refers to a series of processes such as shifting to sleep mode when necessary processing is completed, periodically returning to operating mode (waking up) to perform processing, and then shifting to sleep mode again after processing is completed. It works. By remaining in sleep, deep sleep, or standby modes with low power consumption during periods when the MCU is not processing, the average power consumption can be reduced.

Article header mcu%e3%81%ae%e6%80%a7%e8%83%bd%e6%8c%87%e6%a8%99%e3%81%a8%e3%81%97%e3%81% a6%e3%80%81%e3%81%aa%e3%81%9c%e3%82%a6%e3%82%a7%e3%82%a4%e3%82%af%e3%82%a2% e3%83%83%e3%83%97%e3%82%bf%e3%82%a4%e3%83%a0%e3%81%8c%e9%87%8d%e8%a6%81%e3% 81%aa%e3%81%ae%e3%81%a7%e3%81%99%e3%81%8b%ef%bc%9f1 1

When considering power consumption in operating mode, it is easy to compare only the MCU active current value (vertical axis), but the time axis (horizontal axis) must also be considered.

Article header mcu%e3%81%ae%e6%80%a7%e8%83%bd%e6%8c%87%e6%a8%99%e3%81%a8%e3%81%97%e3%81% a6%e3%80%81%e3%81%aa%e3%81%9c%e5%87%a6%e7%90%86%e6%99%82%e9%96%93%ef%bc%88processing% 20time%ef%bc%89%e3%81%8c%e9%87%8d%e8%a6%81%e3%81%aa%e3%81%ae%e3%81%a7%e3%81%99% e3%81%8b%ef%bc%9f2 1

In other words, the shorter the processing time in operating mode, the lower the average power consumption. How can we shorten the processing time? is the key. How to do efficient software design while making full use of built-in peripherals? is also an important factor, but at the same time, the processing performance of the CPU core itself also occupies a large proportion.
I think that software is often designed in C language, but logic synthesis enables us to achieve what we want by skillfully combining the instruction set of the CPU core. Different CPU cores have different instruction sets, so even if you want to achieve the same thing, there will be CPUs with fewer instructions and CPUs with more instructions. This is directly related to the length of processing time.
Compared to 8/16bit MCUs of the same class in terms of processing time, ARM Cortex-M is a very instruction efficient CPU core.

Article header mcu%e3%81%ae%e6%80%a7%e8%83%bd%e6%8c%87%e6%a8%99%e3%81%a8%e3%81%97%e3%81% a6%e3%80%81%e3%81%aa%e3%81%9c%e5%87%a6%e7%90%86%e6%99%82%e9%96%93%ef%bc%88processing% 20time%ef%bc%89%e3%81%8c%e9%87%8d%e8%a6%81%e3%81%aa%e3%81%ae%e3%81%a7%e3%81%99% e3%81%8b%ef%bc%9f1 1

On the other hand, the larger the number of instructions installed, the larger the die size of the IC and the higher the cost.
EFM32 has variations in the adopted CPU, such as the standard Cortex-M3, the Cortex-M0+ with a slightly reduced number of instructions to reduce the cost, and the Cortex-M4F with improved performance by supporting floating point arithmetic. We offer a well-balanced product line.

Experienced FAE
Free consultation is available.

From specific product specifications to parts selection, the Company FAE will answer your technical concerns free of charge. Please feel free to contact us.