A user integrated a £200 datacenter GPU into a consumer gaming rig. The setup requires custom cooling and specific driver workarounds to function outside a server rack. This experiment proves that older enterprise hardware remains viable for budget AI workloads. Practitioners can leverage these cheap, high-VRAM cards for local LLM inference despite the technical hurdles.