Nscale closes $3 billion in financing for Texas and North Carolina AI sites
- Nscale closes ~$3 billion in senior secured DDTLs for US AI campuses
- Ward County, Texas facility receives up to $1.85 billion commitment
- Madison County, NC facility receives up to $1.2 billion commitment
- Both loans carry investment-grade ratings with stable outlooks
- Funds target GPU infrastructure, networking, and liquid cooling

*this image is generated using AI for illustrative purposes only.
Nscale has closed approximately $3 billion in aggregate commitments across two senior secured delayed draw term loan facilities (DDTLs) to fund its artificial intelligence infrastructure expansions in the United States.
The financing supports the company's Ward County, Texas campus and its Madison County, North Carolina site. Both facilities have received investment-grade ratings with stable outlooks, providing up to $1.85 billion for the Texas location and up to $1.2 billion for the North Carolina site.
Facility Details
The funds will primarily finance the deployment of GPU infrastructure, networking, storage, and liquid-cooling equipment. J.P. Morgan and Goldman Sachs served as joint lead arrangers, joint bookrunners, and co-structuring agents for both transactions.
| Location | Commitment Amount | Key Specifications |
|---|---|---|
| Ward County, Texas | Up to $1.85 billion | Supports ~275 MW IT load; NVIDIA GB300 and VR200 systems |
| Madison County, North Carolina | Up to $1.2 billion | 96-acre site; up to 40 MW capacity; retrofit capex |
J.P. Morgan acted as lead left arranger for the Ward County facility, while Goldman Sachs held that role for the North Carolina facility.
Ward County, Texas
The Texas campus is a purpose-built AI infrastructure site designed for high-density, next-generation compute deployments. It utilizes closed-loop direct liquid cooling and rear-door heat exchangers to support advanced AI systems efficiently at scale.
The loan facility, issued through Nscale Ward County Borrower SPV, LLC, will fund NVIDIA GB300 (Blackwell Ultra) and VR200 (Vera Rubin) systems capable of supporting approximately 275 MW of IT load.
Madison County, North Carolina
The North Carolina site is a 96-acre colocation facility with up to 40 MW of capacity. The associated loan facility will fund site retrofit capital expenditures, GPU infrastructure, and networking required for high-performance AI compute deployments.
What the Numbers Show
The split in financing reflects the differing scale and stage of the two projects. The larger $1.85 billion commitment for Ward County aligns with its significantly higher power capacity (~275 MW vs 40 MW), indicating a focus on greenfield, high-density compute builds. The smaller $1.2 billion allocation for Madison County supports retrofitting an existing colocation site, suggesting a lower capital intensity per megawatt compared to the new build in Texas.
How might the investment-grade ratings for these DDTLs influence the cost of capital and competitive positioning for other mid-tier AI infrastructure developers?
What are the projected timelines for reaching full 275 MW operational capacity at the Ward County site, and how does this align with NVIDIA's supply chain constraints for GB300 systems?
Could the success of Nscale's liquid-cooling implementation in Texas set a new industry standard that pressures traditional air-cooled data center operators to accelerate retrofits?


























