It’s going to be about equivalent to the ATMEGA328P you get on a garden variety Arduino can do, albeit with a lot less IO. You’re probably looking at the speed controller in a power tool or the onboard computer of a Qi charger or something. In a lot of cases, you just need something that can run a few lines of C.
Reading another commenter say it has 16k of flash memory and 1k of RAM in a 2x4 pin package, so even if the core has some speed to it…what exactly are you going to achieve with that?
Other than being 32-bit it’s not too far from the specs of an ATTINY85. I have a couple of those kicking around to play with, it’s 8k of flash, 512 bytes of RAM and 512 bytes of EEPROM and with an external oscillator it can be clocked at up to 20 MHz.
Let’s take my cordless router as an example. It has a brushless motor that can turn at ~20,000 RPM. The microcontroller would have to be able to read the shaft position sensor and make changes to the H-bridge or whatever circuit is driving the motor two if not four times per revolution. How many machine instructions do you think each of those operations takes? Ten? It also has to at least occasionally look at the speed selector to vary the motor RPM, and is probably monitoring the battery or communicating with the battery pack’s own controller to prevent the nine or ten ways you can kill lithium cells by using them.
You’re probably approaching the limit of what you can do with 16k of flash, 1k of memory and 6 IO pins.
I guess maybe we have different requirements, but I built a similar controller using a 68HC05 when I was in college (yes I am that old) and it had plenty of cycles left to burn. This thing is nearly an order of magnitude faster than the microcontroller I used.
EDIT: I will grant you the I/O is pretty limited, but you could also offload some of the work to an external timer/accumulator
I imagine the other factor is the tooling at the chip fabs. When you can make an entire microcontroller that small, for 20 cents apiece, why bother continuing to make less powerful chips? We can just do this now.
Reading the article, they mentioned “medical devices and earbuds” as potential use cases.
It’s going to be about equivalent to the ATMEGA328P you get on a garden variety Arduino can do, albeit with a lot less IO. You’re probably looking at the speed controller in a power tool or the onboard computer of a Qi charger or something. In a lot of cases, you just need something that can run a few lines of C.
A 32bit @24MHz CPU is extremely overpowered for those use cases.
Qi operates in the 87-205 KHz range. A power tool is several orders of magnitude slower.
Reading another commenter say it has 16k of flash memory and 1k of RAM in a 2x4 pin package, so even if the core has some speed to it…what exactly are you going to achieve with that?
Other than being 32-bit it’s not too far from the specs of an ATTINY85. I have a couple of those kicking around to play with, it’s 8k of flash, 512 bytes of RAM and 512 bytes of EEPROM and with an external oscillator it can be clocked at up to 20 MHz.
Let’s take my cordless router as an example. It has a brushless motor that can turn at ~20,000 RPM. The microcontroller would have to be able to read the shaft position sensor and make changes to the H-bridge or whatever circuit is driving the motor two if not four times per revolution. How many machine instructions do you think each of those operations takes? Ten? It also has to at least occasionally look at the speed selector to vary the motor RPM, and is probably monitoring the battery or communicating with the battery pack’s own controller to prevent the nine or ten ways you can kill lithium cells by using them.
You’re probably approaching the limit of what you can do with 16k of flash, 1k of memory and 6 IO pins.
I guess maybe we have different requirements, but I built a similar controller using a 68HC05 when I was in college (yes I am that old) and it had plenty of cycles left to burn. This thing is nearly an order of magnitude faster than the microcontroller I used.
EDIT: I will grant you the I/O is pretty limited, but you could also offload some of the work to an external timer/accumulator
I imagine the other factor is the tooling at the chip fabs. When you can make an entire microcontroller that small, for 20 cents apiece, why bother continuing to make less powerful chips? We can just do this now.
Reading the article, they mentioned “medical devices and earbuds” as potential use cases.