Corvex Signs $103 Million Deal for 1.6 MW of AI Compute Capacity at Its US Data Center

project_news

Corvex Signs $103 Million Deal for 1.6 MW of AI Compute Capacity at Its US Data Center

Updated on Nov 17, 2026, 02:52 AM IST
Written & Edited by Harikesh VA

Corvex, Inc. (Nasdaq: MOVE), an engineering-led AI computing platform specializing in GPU-accelerated infrastructure for AI workloads announced a three-year, USD 103 million take-or-pay agreement to provide NVIDIA Blackwell computing capacity to a scaled AI inference platform.

 

Revenue Set to More Than Double

Under take-or-pay terms, the customer is required to pay for the reserved capacity regardless of utilization. The customer and a lender will fund about 80% of the deployment cost, and the deployment is expected to begin generating revenue by the end of the first quarter of 2027.

Once the deployment is fully in service, Corvex expects contracted take-or-pay annual recurring revenue at its Mid-Atlantic 1 data center to increase from USD 22 million on capacity in service today to approximately USD 56 million, more than doubling the site's contracted annual recurring revenue.

 

Deployment Costs and Financing Structure

The deployment, which includes GPU servers, storage, networking, and spares, is expected to cost approximately USD 78 million.

 

A USD 26 million customer prepayment and USD 38 million in committed debt financing from USD.AI, described by the company as a leading platform financing the Compute Middle Market, together covers USD 64 million, or more than 80% of the cost.

 

Corvex intends to fund the remaining USD 14 million. The debt commitment is a three-year asset-backed loan that is non-recourse to Corvex, Inc.

Trusted by Leading EPCs & Manufacturers

Find the Latest Data Center Projects in the United States

Gain exclusive access to our industry-leading database of data center opportunities with detailed project timelines and stakeholder information

Request Free Trial → Learn More →

No credit card Up-to-date coverage

 

Technical Scope of the Deal

Under the agreement, Corvex will deliver more than 1,000 NVIDIA HGX B200 GPUs connected with NVIDIA Quantum-2 InfiniBand networking.

 

The deployment will use 1.6 megawatts of IT capacity, which the company says equals approximately UDS 21 million of annual contract revenue per megawatt.

Executive Commentary

"This is Corvex's third NVIDIA Blackwell deployment at Mid-Atlantic 1, proving the strength of our supply chain and delivery playbook," said Seth Demsey, Co-Founder and Co-Chief Executive Officer of Corvex. "When fast-growing AI-native companies need certainty, they rely on Corvex to bring the latest generation of high-density compute online quickly and on schedule."

Co-Founder and Co-Chief Executive Officer Jay Crystal framed the deal as a template for the company's growth strategy.

 

"We're diversifying our customer base with another scaled inference platform, and this agreement is a blueprint for how we scale Corvex by securing multi-year take-or-pay revenue alongside non-recourse debt and customer prepayments to expand our footprint efficiently without straining our balance sheet," Crystal said. "This is a repeatable formula for efficient growth."

Facility Readiness and Prior Expansion

Power and space at Mid-Atlantic 1 are on track to support an on-schedule deployment. Corvex announced an expansion of Mid-Atlantic 1's capacity on Aug. 31, 2026.

About Corvex

Corvex, Inc. describes itself as an AI cloud computing company specializing in GPU-accelerated infrastructure for AI workloads and a publicly traded pure-play AI compute platform.

 

The company provides compute resources through GPU-accelerated clusters, high-throughput storage, and a layered architecture engineered for reliability, performance, and efficiency at scale.

 

Its product suite includes AI Factories and GPU Clusters, the Assured AI confidential-computing platform, and the Corvex Token Factory, a serverless inference platform.

Catching the Next Data Center Boom in the United States Starts With the Right Intel

The United States is racing to expand its data center footprint, driven by cloud adoption, AI workloads, and hyperscale investments, but every delayed lead is a contract someone else signed. Developers, EPC firms, equipment suppliers, and service providers all compete for the same narrow window between project announcement and groundbreaking.

 

The challenge isn't a lack of opportunity; it's a lack of early, reliable visibility. Waiting for press releases means arriving late to bids that developers shaped months earlier. What separates winners from spectators is access to verified project intelligence before the market moves.

 

That's exactly what the Global Project Tracking (GPT) platform by Blackridge Research delivers; a continuously updated pipeline of data center projects across the United States, from the first announcement through final commissioning.

 

  • Upcoming Projects

  • Tender Notices

  • Contract Awards

  • Projects Under Construction

  • Completed Projects

 

Ready to see every data center opportunity before your competitors do? Book a Free Demo of the Global Project Tracking (GPT) platform by Blackridge Research today.

Tags

Leave a Comment

We love hearing from our readers and value your feedback. If you have any questions or comments about our content, feel free to leave a comment below.

We read every comment and do our best to respond to them all.

Protected by Cloudflare Turnstile