The $55 million represents capital already invested directly through convertible preferred stock, rather than a newly announced financing round. It excludes additional funds spent by Ault & Company and its affiliates purchasing Hyperscale Data common shares in the open market.
Ault & Company identified Hyperscale Data’s expanding artificial intelligence infrastructure business as a central component of its long-term investment strategy.
The company believes Hyperscale Data’s public market valuation does not adequately reflect the potential value of its underlying assets, particularly its Michigan AI data center campus. That assessment represents Ault & Company’s investment thesis rather than an independently established valuation.
The Michigan facility is owned and operated by Alliance Cloud Services (ACS), a wholly owned subsidiary of Hyperscale Data.
Hyperscale Data has reported investing more than $70 million in the campus, which is being developed to support large-scale AI computing operations.
ACS has executed a master services agreement covering an initial 20 megawatts of critical AI computing capacity. The agreement also gives the customer the right to expand its deployment to 52 megawatts.
According to Hyperscale Data, the initial 20-megawatt deployment could generate more than $1.2 billion in cumulative revenue if the agreement remains in effect throughout its maximum 20-year term.
If the customer exercises its expansion right and the additional capacity remains contracted for the applicable maximum term, total potential contract revenue could exceed $3 billion.
These figures represent potential cumulative revenue over extended contractual periods, not guaranteed revenue or amounts already recognized.
Hyperscale Data estimates that its Michigan campus could eventually accommodate approximately 340 megawatts of total capacity. The contemplated 52-megawatt customer deployment would represent less than 20% of that potential.
However, development beyond the initially contracted capacity remains subject to financing, regulatory approvals, engineering requirements, utility agreements, infrastructure availability, and customer demand.
Hyperscale Data’s management currently estimates that the Michigan campus could support a valuation of approximately $750 million to $1.25 billion.
Management has identified $750 million as the minimum valuation at which it would seriously consider selling the campus. The range is an internal assessment and does not represent an independent appraisal or a formal acquisition offer.
Beyond its AI infrastructure investments, Hyperscale Data reported approximately $360 million in total assets as of June 30, 2026.
The company generated $34.8 million in consolidated revenue during the second quarter of 2026.
Its broader portfolio includes businesses and investments spanning defense and aerospace, crane rental, hotels, commercial lending, blockchain technology, and software.
As of September 13, Hyperscale Data also reported approximately $51 million in Bitcoin, cash, and restricted cash.
Ault & Company said its investment reflects its interest in both the Michigan campus and Hyperscale Data’s established operating businesses.
The holding company intends to continue evaluating additional investments, including its right to purchase up to $96 million in additional Series H convertible preferred stock.
KEY QUOTES:
“We have committed substantial capital to Hyperscale Data because we believe in the value of what it owns and the opportunities ahead. The Michigan campus has an executed customer agreement, significant capital invested and the potential to expand. Management believes that the campus could support a valuation of $750 million to $1.25 billion. Beyond Michigan, Hyperscale Data has established operating businesses and other strategic assets. As its largest stockholder, Ault & Company, together with its affiliates, is focused on supporting Hyperscale in the work needed to develop these assets and demonstrate their value over time.”
Milton “Todd” Ault III, CEO Of Ault & Company And Executive Chairman Of Hyperscale Data
