RISC-V XuanTie C950 CPU Reportedly Runs Qwen 3.8 Without a GPU
TL;DR
Alibaba’s recent announcement about its XuanTie C950 CPU has sparked interest in the AI hardware community. According to The Stack, this RISC-V-based processor reportedly runs the Qwen 3.8 model, a 27-billion-parameter AI system, without relying on a GPU. The CPU achieves a generation speed of 30 tokens per second, with the first token generated in […] The post RISC-V XuanTie C950 CPU Reportedly Runs Qwen 3.8 Without a GPU appeared first on Geeky Gadgets.
Nauti's Take
The direction here is genuinely promising: if a 27 billion parameter model runs at 30 tokens per second on a plain CPU, local inference becomes realistic again for teams without a GPU budget. The catch is the numbers themselves, since they come from Alibaba's own announcement and say nothing about context length, memory footprint or sustained load.
Anyone planning on premise inference should watch this chip and wait for independent benchmarks before budgeting around it.
Video
Sources