Before that, I had an Amazfit Bip for a few years which was incredible. It did notification, GPS, heart rate tracking, always on display and battery life of 2 - 4 weeks. It did this years and years ago, when the best Android could do was 24 - 48 hours, and it did it for like £60 instead of £200. It still works too!

I don't know about Amazfit, but I have a Garmin that also lasts weeks. There are some differences: WearOS/WatchOS watches essentially use a more power-efficient/less powerful version of a smartphone-class SoC. They have to because they run a full Linux/XNU kernel and a pretty complete userland. Watches with weeks-long battery life typically use something that is more akin to a powerful microcontroller with operating systems tailored to such low-end hardware.

Besides that some watches (like several Garmin models) use transflective displays. They do not have to actively emit light during daytime (in contrast to OLED), sunlight is reflected. In contrast, OLED displays have to be more bright in sunlight to be visible.

> Watches with weeks-long battery life typically use something that is more akin to a powerful microcontroller with operating systems tailored to such low-end hardware.

This is what I'd assumed. But then I also assumed that's actually an exceptionally expensive and high resource approach to take compared to using higher level smartphone chips. By using lower level hardware, they're having to do more bespoke hardware design, and more bespoke low-level firmware and software creation, and also support all of that extra creation. This seems like the super expensive, heavy, slow way of building a smartwatch.

So I guess the "what's the deal" what's trying to understand how some random knockoff looking company ("Amazfit" in 2016) took the super expensive, heavy, slow way of building a smartwatch, and got better results than some of the largest most notorious software/hardware companies on the planet.

Ultimately, they took the pebble approach, and pebble also got a huge amount of backing and extra funding, time, support etc. and seemed to commercially fail. But Amazfit is still going strong.