Public equity investors are increasingly receptive to AI data center and neocloud platforms, but issuers are choosing patience over speed as deep private capital pools and structural complexity delay IPO execution.

Investors and advisors in the sector say demand from public equity investors is real. Still, most neocloud operators are not rushing to tap markets, preferring to scale further and simplify structures before listing.

“It seems like an unlimited appetite” for AI infrastructure companies looking to go public, an ECM consultant said. “When you have infinite customer demand – and that’s the case at the moment – you can call the shots.”

An IPO investor tracking the space said sponsors are operating on extended timelines even as appetite from the public side continues to build, with limited listed vehicles offering exposure to AI‑driven data center infrastructure.

CoreWeave is the industry’s shining example. The AI infrastructure firm’s stock price has more than doubled since its initial public offering on the Nasdaq last spring, and its market cap sits above USD 62bn.

Another AI infrastructure company, Corvex, went public via a reverse merger, listing on the Nasdaq last month in an effort to lower its cost of capital and expand access to different pools of financing, co-CEO Jay Crystal told Mergermarket.

Crystal and co-CEO Seth Demsey, who co-founded Corvex in October 2024, have bootstrapped and taken venture capital backing in previous efforts. “This has been, by far, the easiest in terms of accessing capital,” Crystal said, pointing to short fundraising times and light roadshows. “We raised our initial round with one phone call.”

Corvex is one of few AI infrastructure companies that have taken the plunge, in part because private capital remains readily available.

Several neoclouds already generate substantial revenue with metrics that could support a listing, one ECM tech advisor said, but most are not planning to go public until at least next year.

The primary constraint, according to multiple sources, is not valuation or investor interest but structure. Data center platforms often combine stabilized, cash‑generating assets with development‑phase projects that carry higher leverage and longer growth runways. That mix can distort pro‑forma leverage and complicate equity stories for public investors more accustomed to lower‑leverage operating companies, an IPO investor said.

Sponsors have therefore been cautious about coming to market before they can present clearer, more easily digestible structures. This complexity has reinforced the importance of differentiating between legacy data center models and a newer generation of AI‑focused “neocloud” platforms.

Advisors close to the space stress that traditional data center REITs such as Equinix or Digital Realty are not appropriate comparables for AI‑driven operators, which have distinct economics and customer dynamics. AI data centers are increasingly being framed as foundational infrastructure for artificial intelligence rather than simple real estate or capacity providers.

“Data infrastructure is becoming the new software foundation,” the ECM tech advisor said. “As AI systems evolve, data is becoming the most critical layer of the software stack.”

That reframing has accelerated in recent months as neocloud operators evolve into broader platforms. Industry sources describe leading players as moving rapidly from basic GPU compute rental toward vertically integrated AI infrastructure, combining compute, storage, networking, and software layers into full‑stack platforms.

“In many cases, these businesses have scaled at extraordinary speed, moving from tens of millions in revenue to enterprise-scale operations within just a few years,” the ECM tech advisor said.

Despite delays in IPO execution, sources say a lack of public‑market exposure continues to underpin the long‑term equity story. Investors note that scarcity of listed vehicles offering direct access to AI data centers and neocloud infrastructure remains a structural tailwind, particularly as private equity and sovereign investors seek eventual liquidity paths for large‑scale assets.

Many of these assets are growing too big to exit via a sale, several advisors have noted.

Companies to watch

Advisors highlighted Crusoe, Lambda, d-Matrix, and FluidStack as among the likeliest candidates to go public in the near-to-medium-term.

Denver-based Crusoe, which has raised USD 3.9bn in total capital since its founding in 2018, is evaluating another funding round this quarter ahead of a potential IPO, CEO Chase Lochmiller told Mergermarket.

Some view Lambda as a potential leading candidate after hiring IPO bankers, though exact timing remains uncertain as the San Jose, California-based company appointed a new CFO in February.

Santa Clara, California-based d-Matrix, an AI inference hardware startup, was valued at USD 2bn when it raised a USD 275m Series C last November. Earlier this month, it acquired GigaIO’s data center business, expanding its capabilities to support system-level deployments.

London-based FluidStack is lining up, according to one advisor. It was reportedly in talks earlier this year to receive a strategic USD 100m investment from Alphabet valuing the cloud computing firm at USD 7.5bn, as Google seeks to subsidize a network of data-center partners to facilitate broader access to its specialized AI hardware.

And plenty of other players are still raising private capital as they grow.

Together AI, a next-generation cloud infrastructure company, is in discussions to raise “a very large” round of private capital, co-founder and CEO Vipul Ved Prakash told Mergermarket last month. The San Francisco-based company was valued at USD 3.3bn in a USD 305m Series B capital raise in February 2025.

Enterprise grade compute platform Argentum AI is seeking more than USD 100m in capital, founder and CEO Andrew Sobko told Mergermarket in February.

In the meantime, M&A and consolidation play a complementary role. Advisors say private equity‑ and sovereign‑backed roll‑ups remain active as sponsors look to build scale, broaden offerings, and strengthen narratives ahead of eventual public listings.

Platform expansion and vertical integration are increasingly viewed as prerequisites for successful IPOs rather than alternatives to them.

So far, the message is consistent: investor interest is in place, but issuers are not rushing to go public. With private capital still abundant and structural complexity high, neocloud and AI data center IPOs are more likely to emerge in measured waves than as a sudden rush.