Qwen3.6-27 Q8 gets near 100K context on 32GB VRAM

Reddit user shares attempts to run Qwen3.6-27 at Q8 quantization, achieving close to 100K context on a single 32GB GPU. The approach involved aggressive quantization but may not be widely reproducible.