3 ms·
I think it's basically an NPU hanging off a USB connection, so it can be used as a compute device for local LLM inference instead of your existing CPU/GPU.
by majorchord 1mo ago
I think it's basically an NPU hanging off a USB connection, so it can be used as a compute device for local LLM inference instead of your existing CPU/GPU.