Nvidia (NVDA) has moved its Vera CPU from announcement to production, delivering its first systems to some of the biggest names in AI.
Nvidia VP of Hyperscale and HPC Ian Buck personally hand-delivered the first units to Anthropic, OpenAI, Oracle Cloud Infrastructure, and SpaceXAI over two days last week.
NVIDIA’s Ian Buck hand-delivered the first-ever NVIDIA Vera CPUs to our partners @AnthropicAI, @OpenAI, @SpaceX, and @OracleCloud.
Vera is NVIDIA's first custom CPU, purpose-built for the age of agentic AI. This is just the beginning. The road to Vera-powered systems starts… pic.twitter.com/Ep5PLqcqIa
— NVIDIA (@nvidia) May 18, 2026
The first stop was Anthropic’s SoMa offices in San Francisco, where James Bradbury, Anthropic’s head of compute, received the system. Buck brought along a bare Vera CPU motherboard to walk Bradbury through what makes the chip different.
“Scaling compute is an important accelerant for the growth of models,” Bradbury said. “We’re excited to see Vera emerge as a promising part of the ecosystem when solving for agentic workloads.”
The next delivery went to OpenAI’s Mission Bay headquarters, where Sachin Katti, head of compute infrastructure, met Buck on an outdoor balcony. Buck pulled out a screwdriver mid-conversation to pop the lid and show the system’s internals.
The final delivery of the day landed at SpaceXAI’s Palo Alto offices. Nvidia’s team walked Elon Musk through the system’s architecture. Musk asked questions about cores, memory layout, and cooling.
SpaceXAI is evaluating Vera for reinforcement learning and agent-based simulation pipelines in its training stack.
On Monday, the tour continued south to Oracle’s AI Customer Excellence Center in Santa Clara, where OCI product and customer success leadership got hands-on time with the unboxed system.
“OCI plans to deploy hundreds of thousands of Nvidia Vera CPUs beginning in 2026 because agentic AI demands sustained performance at massive scale,” said Karan Batta, who leads overall product management at OCI.
OCI is the first cloud provider to take Vera to hyperscale deployment.
Vera is Nvidia’s first custom CPU, built specifically for agentic AI workloads — the kind where models don’t just answer questions, they take actions, run code, call tools, and manage long-context state.
GPUs handle the heavy compute, but the orchestration layer around them — tool calls, data movement, sandboxing, retrieval — that’s CPU work. That’s what Vera is built for.
The chip packs 88 custom Olympus cores, 1.2 TB/s of memory bandwidth, and delivers 50% faster per-core performance under full load compared to traditional designs.
“When AI models are posed a question, the answer often isn’t already prepped,” Buck said at the Oracle visit. “The models actually have to generate some Python code to arrive at the correct answer. That’s why we are seeing the demand for CPUs skyrocket.”
Vera also serves as the host processor in the Vera Rubin NVL72 configuration, pairing with Rubin GPUs via second-gen NVLink-C2C in a unified memory architecture. Nvidia says the setup runs at 2x the energy efficiency of traditional infrastructure.
Jensen Huang first introduced Vera at GTC San Jose in March, calling it Nvidia’s next multi-billion dollar business.
The post Nvidia (NVDA) Stock: Vera CPU Arrives at Anthropic, OpenAI, Oracle and SpaceXAI appeared first on CoinCentral.
