Altechnative

So, you want a capable home lab for serious AI work, but you don’t want to take out a second mortgage to pay for it?

If you use LLMs on your own hardware, you almost certainly already know that memory needed for the context increases exponentially with the size of the context window.

Public AI Pros: Performance – higher performance than what you can put together yourself on the cheap Cost – free for limited / minimal usage No setup needed Cons: Privacy – never discuss with it anything that you wouldn’t openly post on the internet Usage

Controversial, but demonstrably true.

I’ve recently been trying to put together a remote gaming setup, and while I got it to work with Windows 10, the hardware H.264 encoding was a little flakey – it would drop out all the time at higher resolutions, and only one user could be logged in at the

I recently decided to test effectiveness of LLMs at certain tasks.

A while back my Asrock EPYCD8-2T board started developing weird BMC related problems – boot delays, boot hangs, IPMI being unavailable, even to the point where IMPI interface wouldn’t even ping over the network.

After updating the firmware and maps on the VW I recently acquired, it was time to also bring my Audi’s navigation and infotainment system up to date.

If you are like me and like older vehicles that aren’t permanently internet-connected drive-by-wire computers on wheels for a large array of reasons, you have almost certainly at various times found yourself in the situation where you bought a beautiful e…

Optimizing Memory Consumption on Small Systems – Again Many moons ago, I wrote about optimizing memory availability on a small system with only 512MB of RAM.