Doing anything with your brain will always increase it's aptitude in that "topic".
People that learn lots of human languages become better and better at learnig further human languages.
People that do one type of math get better at learning other types of math.
People that do something with their body coordination (athletes) can pick up other types of body coordination (other sports or dance for example.) faster compared to couch potatoes.
etc.
So doing anything with your brain is already useful, even as just a "workout".
I'm using soquartz compute modules on soquartz blades, because of the nvme slot and PoE. Just one cable for each is so nice in a tiny rack. They are running dietpi and docker swarm with dokploy.
I generally recommend the pine64 stuff, but shipping and tax might be high depending on where you live :/
GLM is pretty good in mg experience, the company I currently freelance at runs it locally (in house server room) for compliance reasons. But it needs very beefy hardware.
From pressing the power on button to an open start menu (I measure like this because the DE is still unusable for a good 5s after it becomes visible) it takes me about 36 seconds.
I recently tried hibernation thinking it would speed things up, but it takes about 2 minutes... huge RAM bad I guess.
(And yes, windows on the same hardware is way faster at about 17s, because it does some magic idk about)
The device operates at a certain wattage. If the voltage is lower, you need more amps to reach the same watt. Amps is what makes conductors hot.
Example:
200V * 1A = 200W
100V * 2A = 200W
To quote the link above that you didn't read so I had to re-read to make sure I'm right:
The issue stems from the difference in standardized voltages between regions, with the US running on 120 volts, compared to China's 220 volts (where bambu is based). This requires almost twice the current for the same total power draw in the US compared to most other regions, contributing to higher temperatures being reached due to Ohm's law.
And about the ac adapter - it's not the cause of the burning issues. And if they specc it for china's 220V, it should work fine in most of europe.
opencode with nanogpt (12$/month subscription or prepaid pay as you go) using GLM 5.1
let me know if you need a referral code for a 5% discount