NVIDIA Unveils Groq 3 LPX for Next-Gen AI Inference

NVIDIA announced the full production of the Groq 3 LPX, extending the Vera Rubin system to enhance agentic AI workloads with faster token generation and superior efficiency.

The Groq 3 LPX aims to optimize token processing for AI agents, improving throughput by up to 30x per watt.

The Vera Rubin rack-scale system integrates NVLink Fusion for better resource utilization.

This development emphasizes NVIDIA's focus on end-to-end AI factory infrastructure.

Source: https://blogs.nvidia.com/blog/vera-rubin-lpx-spectrum-x-nvlink-fusion/