The chipmaker announced on August 24 that the new inference accelerator is now in full production to extend the Vera Rubin NVL72 rack-scale system.
NVIDIA announced on August 24, 2026, that the Groq 3 LPX interactive AI inference accelerator is now in full production.
The company stated the move extends the Vera Rubin NVL72 rack-scale system.
HPCwire reported the same announcement on that date.
Blockchain.News also carried the report that the Groq 3 LPX AI inference accelerator has entered full production on 2026-08-24.
The device is designed specifically for inference tasks.
Read with its sources →