Desktop: NUC - I5 / 16 cores, 64GB ram, 4TB NVME, 2x 1920x1080 monitors, Xubuntu 22.04
Laptop: Framework 13 - I5 / 16 cores, 64GB ram, 1TB NVME, ZenScreen Monitor, Xubuntu 22.04
eGPU: Akito Node Titan / RTX 3060
LLM Runner: Ollama
AI Pair Programmer: Aider
Editor: Neovim
VM Host: Incus
WAN/VPN: Zerotier
Remote Connectivity: Starlink Mini
Tried and Tested: Xubuntu, Neovim, Zerotier
Newer and Awesome: Incus, Ollama, Aider, Starlink Mini
Disappointment: CUDA configuration
Future hardware: Probably something build on a beefy APU for running LLMs. Framework just announced one built on an AMD Ryzen AI Max. 192GB of unified memory!! Apple ships the awesome M3 Ultra (512GB Unified Memory!!) but I will stick with familiar Linux.
Most excited about:
- LLMS.TXT and ExDoc auto-generating markdown - first class LLM support for Elixir packages
- Model Context Protocol - a great fit for Elixir
