LOS ANGELES, May 26, 2026 — INNIO is not showing up as just another industrial IPO. The company sits in a part of the power stack that public investors care about right now: fast-start distributed generation for data centers, grid support and backup power. That gives the deal real thematic heat. But the amended S-1/A filed May 26 also makes the basic trade-off unusually plain: this is a roughly $2.0 billion IPO at the midpoint, and the money is going to the selling shareholder, not into the business.

The terms are large enough to matter. INNIO is marketing 75 million shares at $24 to $27 each, implying gross proceeds of about $2.025 billion at the midpoint, with a standard 11.25 million-share greenshoe. The company has applied to list on Nasdaq as INIO, and the syndicate is as deep as the calendar gets, led by Goldman Sachs, J.P. Morgan and Morgan Stanley. IPOGrid reads that combination as important: the deal has scale, real sponsorship and serious banking support. It also has the feel of a monetization event first and a capital-raising event second.

That tension exists because the operating story is not fake. In the public S-1 filed May 11, INNIO said annual data-center equipment order intake climbed from $27 million in 2023 to $2.282 billion in 2025. The company backed that narrative with unusually tangible issuer-side newsflow: an October 2025 VoltaGrid order for 2.3 gigawatts, a February 2026 follow-on VoltaGrid order for 1.5 gigawatts, an April 2026 framework agreement with Rehlko for 1.25 gigawatts of gas-engine capacity, and an April 2026 hydrogen backup-power demonstration for data centers witnessed by Microsoft, Google and Data4. For IPO readers, that is better evidence than a generic AI-adjacent slide deck.

The financial profile is solid enough to keep the story from collapsing into pure concept. In an earlier amended draft filing, INNIO disclosed $3.884 billion of 2025 equipment order intake, $2.6368 billion of revenue, $549.0 million of adjusted EBITDA and $141.8 million of net income for 2025. The same filing showed order intake split heavily toward power solutions, with data centers taking a much larger share of the mix than in prior years. The public S-1 also points investors to a recurring services engine, with $1.2714 billion of 2025 services revenue and a total revenue base that rose from $2.015 billion in 2023 to $2.1591 billion in 2024 and $2.6368 billion in 2025. That is a more credible setup than many energy-transition IPOs that arrive with little operating proof.

Still, the reviewer’s concern is that investors are being asked to pay for momentum while accepting a structure that leaves the issuer itself on the sidelines. INNIO stated in an amended draft prospectus that it will not receive any proceeds from the sale of common shares by the selling shareholders. The public filing also says AI Alpine (Luxembourg) S.à r.l. will remain the principal shareholder and INNIO will qualify as a controlled company after the offering. That does not make the deal broken. It does mean public buyers are underwriting liquidity for existing owners while getting minority exposure to a sponsor-controlled company.

The near-term numbers also argue against treating this as a no-questions-asked AI power trade. Bloomberg reported from the May 11 filing that INNIO posted a net loss of $7.2 million on $668.6 million of revenue in the first quarter of 2026, versus net income of $35 million on $494 million of revenue a year earlier. The filing record also shows $841.2 million of cash and cash equivalents alongside a leveraged capital structure that included a $750 million Term Loan B2 established in October 2025 and a February 2026 amend-and-extend transaction that pushed major term-loan maturities to November 2031. Add in the material weaknesses in internal control over financial reporting disclosed in draft filings, and the picture gets more complicated than the headline growth would suggest.

Our interpretation is that INNIO deserves attention because both halves of the story are real. There is genuine demand color here: data-center orders, large named counterparties, a top-tier bank group and enough scale to matter. There is also a clear public-market ask: buy into an energy-infrastructure supplier with improving historical results, uneven recent profitability, sponsor control and no primary proceeds. If the book builds anyway, that will say something useful about how much investors are willing to pay for AI-power exposure even when the issuer is not the one getting the cash.