Search results for “LPU” (1)

G

Groq runs open models on its custom LPU chips for extremely fast inference, hundreds of tokens per second, ideal for la…