Show Your Setup: What's Running Your Local AI?

I'll go first. AMD Ryzen 7 7840U (Phoenix1 APU), 24GB DDR5, Ubuntu 24.04, Vulkan. Running 40 Python services + Qwen 3B

• WebUI concurrently. Remote RTX 5070 for LoRA training via Tailscale SSH. Total cost: ~800 EUR laptop + borrowed training rig + 50 EUR/month electricity.

What I'd upgrade: more VRAM. 8GB shared is the ceiling for 3B.

Show me what you're running. Especially unconventional setups.