Strix Halo makes 122B local inference usable

This is the kind of benchmark that makes local AI hardware feel materially different, not just incrementally faster. Strix Halo is starting to look like a legitimate single-box inference platform for developers who want to run huge GGUF models without jumping to a discrete GPU rig.

• –The real headline is not the tiny-model speed, but that a 122B-class quant is actually usable on integrated graphics with unified memory.
• –MoE models look especially strong here, with Qwen3.5-35B-A3B and GLM-4.7-Flash posting numbers that make local experimentation much more practical.
• –ROCm 7.2 appears to be a meaningful step for AMD's local LLM story, especially versus the usual perception that CUDA is the only serious path.
• –For AI developers, this points to a new sweet spot between laptop-class convenience and workstation-class local inference capacity.
• –The benchmark is still a community Reddit post, not a vendor-run validation, but the results line up with the growing interest in Strix Halo as a local-LLM box.