

Same, picked up an Asrock x300 and running a 4750g at 35w.
Perfect for anything the house needs, and use the gaming pc for testing and prototyping


Same, picked up an Asrock x300 and running a 4750g at 35w.
Perfect for anything the house needs, and use the gaming pc for testing and prototyping
You don’t even get a cubicle anymore, here’s a lopsided ikea desk you share with three other people.


Suck my dick O’Leary


China parity with TSMC is always a year or two away
https://www.infoworld.com/article/2217963/china-s-smic-settles-lawsuits-with-tsmc.html


It’s almost as if nuance and context matters.
How much energy does a human use to write a Wikipedia article? Now also measure the accuracy and completeness of the article.
Now do the same for AI.
Objective metrics are what is missing, because much of what we hear is “phd-level inference” and it’s still just a statistical, probabilistic generator.
https://www.pcmag.com/news/with-gpt-5-openai-promises-access-to-phd-level-ai-expertise


lol, I have llama.cpp and ollama setup on a separate pc just so I can understand and discuss from experience.


Tangible value, not just a black box that might give you what you want
Bypassing the GFW is illegal
Yes, I do know and realize that. Why it’s probably not a good idea to try connecting to your homelab lol
What you’re asking is illegal where you’re going
Best of luck to you


Figure head syndrome. Literally a man with money, otherwise completely unremarkable


Cloudflare isn’t very self-host, unless you want/need to trust a third party I wouldn’t recommend this.


Good news! Emojis are also available in GMail now!
https://mspoweruser.com/google-workspace-adds-emoji-api-expands-gemini-features-in-latest-update/


Reddit, created by users and Russian bots
When support is gone, time to port if possible


Ahh yes, optane redux


Looks pretty interesting, thanks for sharing it
Facism, welcome to 2008