• 2 Posts
  • 116 Comments
Joined 5 months ago
cake
Cake day: December 24th, 2025

help-circle









  • It is completely up to your local market and needs, ie for how much you can sell it, what you can buy etc…

    The only tips I can give you:

    • Do you REALLY need all of the compute power? Most people don’t and you would wast a ton on electricity. This point on it’s own is enough to push most to a mini pc. (Immich can run on basically any mini pc)
    • is there absolutely no other place for the server where the noise wouldn’t be an issue?
    • look into sound proofing, you could house it in a box lined with acoustic foam with a the airway taking multiple 90° turns to keep airflow while reducing noise.

    P.S:

    I personally went on an in-between, I have a large tower pc, basically a server but with hardware meant for mostly silent work, so it rarely get’s noisier than background.





  • Phone manufacturers don’t release drivers, leaving phones in a state of either it works with generic drivers (almost never works 100%) or someone needs to make a custom driver (a lot of work) so we end up in a situation where every new phone would need a full team working on it for weeks/months just to make Linux support it, so it rarely happens. and when it does, by the time you reach full functionality, the phone is already outdated.

    With computers this doesn’t happen as they are essentially modular (being built from mostly a combination of off the shelf parts) where each part already has a driver (often even contributed by tge manufacturer) while phones are almot completely custom, with each model having custom parts that are often completely unique to it.




  • Wthout going into the issue itself, it is such a ridiculous waste to use an llm for something that a far simpler model could do like 100x faster and locally for essentially free…

    Just search for “machine learning text moderation” and you will find all kinds of options. Not to talk about the fact that a simple 4B LLM could do this as well.

    One thing I really hate is how LLMs have completely overshadowed the entire ML/AI field and people just use them for everything.

    Using a trillion parameter LLM model for basic text moderation is like using a gaming rig to play candy crush.