NVIDIA Vera Rubin Platform Ships Seven Chips in Full Production; 10x Inference Cost Reduction Over Blackwell

NVIDIA announced seven Vera Rubin platform chips are now in full production, including the Vera CPU, Rubin GPU, NVLink 6 Switch, ConnectX-9 SuperNIC, BlueField-4 DPU, Spectrum-6 Ethernet switch, and Groq 3 LPU. The platform delivers up to 10x reduction in inference token cost and 4x reduction in GPUs needed for MoE model training versus Blackwell. AWS, Google Cloud, Microsoft, and OCI will be among the first to deploy Rubin-based instances in 2H 2026. NVIDIA CEO Jensen Huang said OpenClaw was 'probably the single most important release of software ever,' drawing a direct line between infrastructure scaling and agentic AI adoption.