Neocloud said in its earnings report that over $4 billion in annual recurring revenue projected for 2026 has been delayed to at least January, with the possibility of extending into February. The company identified Microsoft as the customer requesting the change, citing a need to "reconfigure and upgrade some of the network infrastructure" to support the latest generation of AI models. Co-CEO Daniel Roberts explained that the work was designed to enable 200 megawatts of capacity to service these systems.
Despite the revenue delay, Neocloud reported strong quarterly results. Revenue climbed 66% from the prior quarter to $228 million, exceeding the $221 million average analyst estimate according to LSEG data. AI Cloud revenue more than doubled to $200 million, while the remainder came from bitcoin mining, the company's original business before pivoting to AI infrastructure. The company's net loss narrowed to $263 million from $684 million in the previous period.
The revenue pushback reflects broader caution in the AI infrastructure sector following recent market turbulence. Neocloud's stock price fell 26% after the recent election, with shares down 55% over the past month before rising in extended trading following the earnings announcement. Other AI infrastructure companies have fared worse: CoreWeave has plunged 60% in the past month, while Nebius is down 48% over the same stretch.
Roberts told CNBC that "capital markets have become more selective in the last few weeks, and the questions have undoubtedly moved from how big the contract is to whether it can be delivered, financed, and run at a profit." However, he also noted a disconnect, saying "There is a disconnect between what we're seeing in public market and what we're seeing at the customer level." Neocloud said the $4 billion in delayed revenue is fully contracted and funded.
Looking ahead, Neocloud plans $15 billion in capital expenditures for the rest of this fiscal year, comprising $9 billion in cash and committed financing plus $6 billion in "target" financing subject to market conditions. The company said revenue backlog increased 22% to $20.3 billion, with approximately two-thirds with investment-grade counterparties. Roberts noted that Nvidia and Microsoft would back some data center construction, while about a third would be funded through large customer prepayments, with one customer agreeing to fund 90% of GPU capital expenditures.
For savers and investors in the US, this signals that while AI infrastructure demand from major tech customers remains strong and contracted, execution risks and financing pressures are mounting as broader market conditions tighten.