Notes for April 8-14

In short, I spent a fair chunk of my time dabbling with LLMs again, but also still dealing with shifting priorities at work.

As noted in my update on the am18, iGPU compute is still fraught with issues–and the apparent bottleneck in data transfer between CPU and GPU plus fragmentation when reserving memory doesn’t help, so I’ll be going back to SBCs for a little while.

Other than that, progress on my projects has been much slower–I did redesign and print a case for a second custom PiKVM, but it’s been quite hard to find enough free time to get interesting stuff done.