TL;DR — Key Takeaways
- Anthropic has reportedly committed $35 billion to NVIDIA-backed Lambda for AI computing capacity at a Texas data center.
- NVIDIA is expected to hold the lease on the facility, further deepening its role in the AI infrastructure ecosystem beyond chip sales.
- The deal follows Anthropic’s $45 billion agreement with Nscale, underscoring how aggressively frontier AI companies are securing long-term compute capacity.
Anthropic has signed a $35 billion cloud computing deal with NVIDIA-backed neocloud Lambda, and NVIDIA will reportedly hold the lease on the Texas data center providing computing capacity for Anthropic’s AI operations.
The agreement calls for Lambda to provide Anthropic with capacity from a data center in Nueces County, Texas. Hut 8, a data center operator that previously focused on cryptocurrency mining, is developing the project, which will provide about 350 megawatts of capacity.
The deal further expands NVIDIA’s leading role in the infrastructure buildout supporting AI, as it invests in cloud providers and other businesses that help create demand for its processors.
Lambda is one of those providers. NVIDIA has invested in the AI cloud company, which specializes in providing GPU-based computing infrastructure. By supporting companies like Lambda, NVIDIA helps expand the pool of available AI computing capacity while creating a larger market for its own hardware.
“That is the second NVIDIA-backed provider Anthropic has committed to this month, which puts the chip supplier between the model vendor and its own capacity,” said Mitch Ashley, VP and Practice Lead at The Futurum Group.
“Platform teams standardizing on a frontier model inherit that arrangement whether they priced it or not. It is important to ask where your vendor’s capacity is contracted, who owns the site, and what your roadmap is if financing or business conditions tighten.”
Anthropic’s Aggressive Expansion
For Anthropic, the $35 billion commitment is another step in an aggressive expansion of its computing resources. The Claude developer needs vast amounts of processing capacity both to build new AI models and to operate products like its Claude Code development tool.
Just last week, Anthropic agreed to spend $45 billion to rent AI computing capacity from Nscale at a data center campus in West Virginia. It has also reached other major cloud agreements as it seeks to secure infrastructure for future demand.
The scale of these commitments highlights how competition among leading AI companies is moving beyond individual models. Access to GPUs, data centers and sufficient electrical power are becoming competitive requirements. AI companies that cannot obtain enough computing capacity face limits on the number of customers they can serve.
Lambda is positioning itself to benefit from this demand. The company raised more than $1.5 billion in November and is reportedly discussing a new funding round of as much as $3 billion, potentially at a valuation of $12 billion or more.
Lambda also reached an agreement with Microsoft last year to deploy AI infrastructure using tens of thousands of Nvidia processors. Adding Anthropic as a major customer will give Lambda another prominent role in the expansion of GPU-based AI infrastructure.
Hut 8 is making a similar transition toward the AI market. In July, the company announced a second 15-year, $9.8 billion lease with an unnamed investment-grade customer, bringing the base-term contract value for the two leases at its Beacon Point campus to $19.6 billion. NVIDIA was later reported to be the tenant.
For NVIDIA, these relationships create opportunities beyond direct chip sales. Its financial agreements help cloud companies and data center operators bring additional GPU capacity online, which enables the chipmaker to play a lead role in developing the infrastructure for AI in addition to selling the GPUs that drive it.
