Nvidia Corp. plans to lease a major data center in Nevada backed by a $3.8 billion junk bond sale, the latest sign of a surge in borrowing related to artificial intelligence infrastructure.
A company backed by alternative asset manager Tract Capital plans to issue high-yield bonds to partially finance construction of a 200-megawatt data center and power substation in Story County, Nevada, according to people familiar with the matter. According to reports, the size of the bond issue has been increased by $150 million following strong investor demand, with pricing discussions suggesting a yield of around 6%.
The planned facility highlights how developers are increasingly leveraging high-yield markets to fund AI-driven projects. Investment-grade borrowers like Oracle and Meta Platforms have raised billions of dollars to expand data capacity, and now more speculative issuers are following suit to capitalize on the AI momentum.
Nvidia will begin the first lease under the new agreement, which will last 16 years, but the company can choose to extend the lease two more times for 10 years each. Companies developing advanced AI chips and systems need to expand their computing resources as model development and chip design tasks become more complex and require more power.
Tract Capital, which manages approximately $6 billion in assets focused on digital infrastructure, has introduced a fleet strategy aimed at building new facilities exclusively for hyperscale customers.
The Fleet I Fund serves as the first fund for companies that own the borrowing entities that issued the bond sales. The fund will provide approximately $620 million in capital for the project, but the additional debt proceeds will reduce its equity share.
JPMorgan Chase & Co., which is leading the deal along with Morgan Stanley and other banks, is also participating in the deal, which is expected to reach pricing soon.
While Nvidia operates its own data centers, it leases additional data center space from major cloud service providers such as Amazon Web Services and Microsoft Azure as its AI training and deployment needs continue to grow.
