Gartner said on 14 October that more than $3.5 trillion of AI infrastructure spending could lose value by the end of 2030. That holds if AI providers keep spending at their current pace. Gartner expects them to spend nearly $1 trillion in 2026 and another $6 trillion from 2027 to 2030. It also predicts that over 80% of neoclouds, the newer firms that rent out AI computing power, will stop operating independently by 2030. The Decoder, citing The Information, reported on 19 October that Firmus pulled its IPO. It said shares of CoreWeave and smaller rivals fell about 20 percent in a week.
Gartner says AI data center owners would need $8.5 trillion in revenue through 2030 to cover their costs and a 12% return. It forecasts no more than $3.6 trillion. These are forecasts, not results, and the full Gartner report is open to clients only. Neither source says which neoclouds will fail or when. The Telegraph reported that Microsoft pulled out of a planned data center deal in Loughton, England, run by Nscale. Nscale said its relationship with Microsoft remains strong.
Anyone who rents AI computing from a neocloud is relying on a supplier Gartner expects mostly to stop operating on its own. The Decoder adds that customers outside the largest cloud companies, neoclouds among them, brought in $40 billion for Nvidia in the July quarter, 45 percent of its data center revenue.