The data-center IPO market is becoming a test of execution, not just AI ambition. Higher interest rates raise the stakes for capital-intensive projects, while investors want clearer evidence that planned capacity can be powered, completed and converted into revenue. Singapore-based operator DayOne is pursuing a possible November U.S. listing; SoftBank-backed SB Energy has delayed its offering amid questions about its reported valuation and reliance on a small group of customers.
1
2
Two IPO plans, different evidence for investors
DayOne is reportedly considering a U.S. IPO as soon as November, potentially raising up to $5 billion at a valuation of about $20 billion. Those figures are reported expectations, not finalized offering terms.
1
15 DayOne is described as a data-center operator, but the available reporting does not provide enough comparable operating and financial detail to assess its capacity, power supply or customer mix against SB Energy’s figures.
1
SB Energy’s proposed valuation has been reported at $50 billion or more, but the company’s data-center business had no operational data centers when it filed for an IPO. Its reported $439 billion backlog and 8.8 gigawatts of data-center capacity refer to contracted or under-construction projects, not facilities already in service.
2
5
6
That difference matters. A large backlog signals expected future business, but it is not the same as current data-center revenue. SB Energy reported about $139 million in first-half 2026 sales from legacy energy operations, with no revenue from operating data centers. Its near-term data-center leases are tied to SoftBank and OpenAI, concentrating customer exposure.
4
5
Why SB Energy’s offering has faced resistance
A valuation above $50 billion asks investors to look ahead to projects that still need to be built and brought online. The reported customer concentration adds another risk: long-term leases provide evidence of demand, but dependence on a few customers can make that demand less diversified.
2
4
5
The company’s figures should therefore be read together, not in isolation. Its contracted capacity and backlog describe potential future activity; its lack of operating data centers and data-center revenue describes the distance still to travel before that activity becomes realized business.
4
5
6 Investors have questioned the valuation, and reports say bankers struggled to find enough buyers at the sought-after price range.
2
What DayOne’s timing does—and doesn’t—tell us
DayOne’s decision to keep moving toward a possible November listing shows that some data-center IPO plans are still advancing, even as higher rates and setbacks elsewhere threaten to narrow the fundraising window.
1 It does not establish that its proposed timing or valuation is secure: both remain reported expectations rather than settled terms.
1
15
Nor do the available reports support a full like-for-like comparison of DayOne and SB Energy on operating capacity, secured power or revenue visibility. The clearest distinction in the reported figures is that SB Energy’s data-center portfolio was not yet operational, while DayOne is described as an operator.
1
5 Investors will need company disclosures to judge how much capacity is already serving customers, how much is still planned, and what power and contracts support it.
Power and construction risk reach beyond one IPO
A project’s customer commitments do not remove the risk that construction or power delivery will fall behind schedule. Oracle issued a force-majeure notice related to Blue Owl’s New Mexico data-center project, citing potential power delays; Reuters reported the project could face about a one-year delay.
19
20 The episode underscores why investors examine not just contracted demand but also whether the infrastructure needed to serve it can be delivered on time.
Switch, Vantage Data Centers and CyrusOne have also been reported as potential IPO candidates.
13 Their presence points to a pipeline of possible listings, not proof that the market will welcome every offering. For each company, the key questions remain practical: what is operating, what is still under construction, how secure is the power supply, who are the customers, and when can those commitments produce revenue?
The market’s dividing line is increasingly between capacity that is contracted and ready to deliver, and capacity that remains a promise on a project schedule. DayOne’s progress and SB Energy’s delay show why an AI infrastructure story alone may not be enough to satisfy public-market investors.
1
2
30