Anthropic, the firm behind the Claude chatbot, has finalized a $35 billion agreement with Lambda, a cloud infrastructure specialist backed by Nvidia. This partnership underscores the intensifying race for the hardware, electricity, and data center capacity required to sustain advanced artificial intelligence.
Texas Infrastructure and Nvidia's Strategic Role
The project centers on a data center in Nueces County, Texas, currently under construction by the infrastructure firm Hut 8. According to sources familiar with the matter, Nvidia is slated to lease the facility, further cementing its role as a primary supporter of Lambda.
This arrangement highlights Nvidia's evolving market presence. Beyond supplying processors, the company is increasingly involved in the financing, leasing, and development of the physical sites that house AI systems. This model creates a cycle of interdependence: AI firms require compute power, cloud providers need capital, and Nvidia provides both, effectively fueling demand for its own hardware.
A $175 Billion Commitment Portfolio
The Lambda contract is part of a broader $175 billion commitment strategy by Anthropic to support its model development. Recent multi-year deals include:
- $45 billion with Nscale for capacity in West Virginia.
- $50 billion with Fluidstack.
- $45 billion with Elon Musk’s SpaceX.
- $35 billion with Lambda.
While these represent long-term contractual obligations rather than immediate annual spending, the cumulative total demonstrates the staggering costs of training and operating advanced AI. These systems require not only more processors but also vast amounts of electricity and specialized cooling networks.
Lambda’s Market Ascension
For Lambda, this agreement solidifies its standing in the specialized AI cloud market. The company is currently in talks to raise up to $3 billion in new capital, potentially reaching a valuation of $12 billion or higher. Having already secured a deal with Microsoft to deploy tens of thousands of Nvidia GPUs, the Anthropic contract provides Lambda with predictable future revenue to offset the massive upfront costs of building and equipping new data centers.
