News

Nvidia adds a seventh processor to its upcoming Vera Rubin AI platform

Tuesday, March 17, 2026 at 06:58 PM

Nvidia has announced an expansion of its upcoming Vera Rubin architecture, adding a seventh processor to the previously announced six-chip platform. This platform represents the next generation of Nvidia's AI infrastructure, following the Blackwell series.

Context

At the GTC 2026 conference on March 16, 2026, Nvidia expanded its upcoming Vera Rubin AI platform by adding a seventh specialized processor, moving beyond the six-chip architecture originally detailed at CES 2026. This updated platform, designed for the era of agentic AI and reinforcement learning, now includes the Vera CPU, Rubin GPU, NVLink 6 Switch, ConnectX-9 SuperNIC, BlueField-4 DPU, and Spectrum-6 Ethernet Switch, alongside new high-performance storage or inference-specific accelerators like the Groq 3 LPU. This integration allows the entire rack to function as a single unit of compute, achieving a 10x reduction in inference token costs compared to the previous Blackwell generation. The Vera Rubin platform is scheduled for full production and partner availability in the second half of 2026, with major cloud providers like AWS, Google Cloud, and Microsoft set to deploy instances. Key hardware highlights include the 88-core Vera CPU, which delivers 1.2 TB/s of memory bandwidth, and the NVL72 rack-scale system that provides up to 50 petaFLOPS of FP4 performance. This aggressive roadmap aims to support the transition to million-GPU environments and orbital data centers, with Nvidia claiming the new architecture is 50% faster and twice as efficient as traditional rack-scale solutions.

Related Companies

Nvidia
Nvidia
NVDA
US