Tag
A user reports that their Asus Ascent with Nvidia GB10 (DGX) is slower than their Ryzen AI Max when running LLMs like Gemma4-31B, despite expected 2-4x speedup, and shares their llama-cpp configuration for debugging.
A user reports successfully using tap water to cool a DGX server while running the Qwen3.5-122b model at high GPU utilization, maintaining safe temperatures.