Loading memory…
The Groq 3 LPU was integrated via NVIDIA's licensing deal with Groq to provide low-latency inference processors.