CPU 推理CPU Inference
CPU Inference
解决什么问题What problem it solves
没有显卡也能跑模型。You can run models without a GPU; the cost is speed.
这是什么What it is
llama.cpp 官方称其目标。概括是「以最少配置、在广泛硬件上实现 LLM（与 VLM）推理」。
官方列出 BLAS、ZenDNN、OpenVINO 等面向 CPU 的后端。
llama.cpp states its goal is LLM (and VLM) inference with minimal setup and state-of-the-art performance on a wide range of hardware.
It lists CPU-facing backends such as BLAS, ZenDNN and OpenVINO.
为什么需要知道Why it matters
决定「没有显卡能不能玩」。能不能跑、跑多快，取决于后端与硬件。Answers "can I run this without a graphics card". Yes, but expect tokens per second, not per millisecond.
现在能做什么What you can do now
回到名词地图看第5站的其他 29 个词 →Back to the map: the other 29 terms in station 5 →
出处Sources
- llama.cpp 官方仓库llama.cpp repository ↗