Fermi Inc. Accelerates AI Power Infrastructure with 2 GW Secured and Strategic Evolution

Fermi has used this update to make one big point: it is no longer pitching itself as a scrappy concept story. Management wants investors to view it as a serious infrastructure business building private power for the AI boom, with real kit, real permits and real financing behind it.

That is the bullish read. The more cautious read is just as important: Fermi is still pre-commercial, still loss-making, and still needs to convert interest from customers into a binding tenant agreement. Until that happens, this remains a high-potential project with meaningful execution risk.

Fermi Q1 2026 results: the key numbers retail investors should focus on

Metric	Q1 2026 / March 31 2026	Comment
Total cash and restricted cash	$243.3 million	Liquidity remains sizeable, but lower than year-end 2025
Cash and cash equivalents	$207.5 million	Down from $408.5 million at December 31 2025
Restricted cash	$35.8 million	Cash set aside for specific purposes
New equipment finance facilities secured	$785 million	Includes a $500 million facility from MUFG
Yorkville financing commitment	$156 million	For general corporate purposes
Outstanding debt	$421.3 million	Up sharply as financing ramps up
Net loss	$188.7 million	Equal to $0.30 per diluted share
Share-based compensation	$134.0 million	Non-cash, but very material
Loss on extinguishment of debt	$24.8 million	Linked to repaying the Macquarie term loan
Capital invested in property, plant and equipment in the quarter	$441.2 million	A huge build-out push
Property, plant and equipment, net	$1.43 billion	Up from $935.3 million at year-end 2025

Why Fermi 2.0 matters for the AI power infrastructure story

Management is rebranding this phase as “Fermi 2.0”. Strip out the corporate language and the message is simple: the company thinks the asset base is now substantial enough that governance, leadership and commercial discipline matter just as much as construction speed.

That matters because AI infrastructure investors care about two things above all else – power availability and credibility. Fermi says it has already built or secured more than $1.4 billion of infrastructure at Project Matador, and now wants to prove it can operate like an institutional-grade public company rather than a founder-led startup.

On the positive side, that is sensible. Big counterparties such as hyperscalers – giant cloud and computing companies – do not just buy capacity from a good idea. They want robust governance, clear financing, experienced executives and confidence that a project can actually deliver power on schedule.

On the negative side, the company is effectively admitting it is in transition. A CEO search is still underway, there is an interim CFO, and the next 90 days are framed around securing a binding tenant agreement. That tells you the commercial piece is not finished yet.

Project Matador progress: 2 GW secured today, 6 GW permitted, 17 GW ambition later

This is the operational heart of the announcement. Fermi says it has secured more than 2 GW of total power generation across owned and contracted assets, giving it a clear path to commercial power delivery later this year.

For investors, that is the most tangible milestone in the whole release. Power is the product here. If Fermi can start delivering it on-site, the story shifts from future promise to early execution.

The permitting update is also important, but it needs reading carefully. The company has obtained an approximately 6 GW Clean Air Permit from the Texas Commission on Environmental Quality, and has filed for an additional approximately 5 GW permit.

So yes, it is advancing toward around 11 GW of permitted capacity, but only 6 GW is actually permitted today based on this RNS. The extra approximately 5 GW is an application, not a done deal. And the full 17 GW build-out remains a long-term ambition, subject to further approvals, financing, interconnection and other factors.

There was more real-world progress too: nearly 5 miles of natural gas lines completed, more than 11 miles of perimeter fencing installed, over 7 miles of on-site water distribution lines finished, and both transmission systems connected for later grid interconnection. The first six Siemens SGT-800 gas turbines have also arrived in the Port of Houston and cleared customs.

Those details matter because they show this is not just a permitting narrative. Physical infrastructure is being put in the ground. In capital-heavy energy projects, that is often where the line between promotional talk and genuine progress starts to appear.

Commercial momentum sounds better – but non-binding is still non-binding

Fermi says tenant engagement has improved in recent weeks and that it is in active discussions with hyperscalers, neo-cloud providers and enterprise compute operators. It also says recent site visits from prospective tenants and strategic partners have increased urgency in commercial conversations.

That is encouraging, especially because AI data centre demand is currently obsessed with near-term power access. Fermi is clearly trying to position Project Matador as a rare asset that can offer scale faster than rivals.

But investors should not over-read this. The company explicitly talks about reaching new non-binding long-term agreements in the near future, while separately saying one of its next-90-day goals is securing a binding tenant agreement. In plain English, interest is up, but signed revenue-backed contracts are not disclosed yet.

That is the biggest gap in the story. There is no revenue line in the income statement, and the business remains pre-commercial based on the information provided.

Fermi financial results: heavy losses, but the quality of the loss matters

The headline net loss of $188.7 million looks ugly at first glance, and it should not be ignored. However, the detail matters here.

The company says the loss was primarily driven by $134.0 million of non-cash share-based compensation and a $25.0 million extinguishment loss on the Macquarie term loan. Non-cash means it hits reported profit, but not immediate cash in the same way as wages or supplier payments. That makes the accounting loss less alarming than the headline suggests, though still significant.

Cash flow tells a more useful story. Net cash used in operating activities was $7.3 million, while investing activities used $441.2 million. So most of the cash outflow came from building the project rather than day-to-day operations.

That is actually what you would expect from a company in heavy construction mode. The trade-off is obvious: if commercial agreements slip, investors may start worrying that large capital spend is getting too far ahead of contracted demand.

Balance sheet, debt and liquidity: enough firepower for now, but leverage is rising

Fermi ended the quarter with $243.3 million of total cash and restricted cash. That is a meaningful liquidity position, and management has added serious financing support with $785 million of equipment finance facilities plus the $156 million Yorkville commitment.

That said, outstanding debt rose to $421.3 million from $109.8 million at the end of 2025. This is not automatically bad – project businesses often use debt once assets are being deployed – but it does raise the stakes on execution.

In short, Fermi has funding momentum, but it also now has a more leveraged balance sheet. If it lands a strong tenant agreement, that leverage could look smart. If not, the market may become much less patient.

Governance changes at Fermi: sensible upgrade before the next growth phase

The governance changes look constructive. The board has expanded from five to seven directors, Marius Haas has become chairman, Robert Masson has joined as interim CFO, and Heidrick & Struggles has been hired to find the next CEO.

For a retail investor, this is one of those updates that can seem boring but is actually useful. Large infrastructure customers and lenders prefer dealing with companies that have proper board oversight, finance leadership and succession planning. It lowers execution risk, even if only a bit.

What this Fermi RNS means for shareholders

My view is that this is a positive update overall, but not a clean slam dunk. The positives are substantial: more than 2 GW secured, a 6 GW permit in hand, more financing lined up, physical construction advancing, and signs that commercial interest is improving.

The negatives are just as real: no binding tenant agreement disclosed, no revenue disclosed, a large quarterly loss, rising debt and a business model that still depends on permits, interconnection, financing and customer conversion all landing in roughly the right order.

So the investment case is becoming more tangible, but it is still not de-risked. If Fermi delivers a binding tenant agreement and starts commercial power delivery later this year, this RNS could end up looking like an important turning point. If those milestones slip, investors will probably focus less on the ambition and more on the cash burn and leverage.

For now, Fermi looks like a company moving from story stock to execution stock. That is progress. It also means the market will judge it more harshly from here.