Neoclouds Race for the Public Markets, Leaving VCs Behind
Deep Dives
Explore related topics with these Wikipedia articles, rewritten for enjoyable reading:
-
Fabless manufacturing
12 min read
Understanding the fabless model is essential context for why neoclouds can exist - they build on infrastructure without owning chip fabrication, similar to how Nvidia designs but doesn't manufacture chips. This explains the industry structure enabling 'sweetheart relationships with chipmakers'.
-
Initial public offering
1 min read
The article centers on neoclouds racing to go public, which is described as 'unusual in the current tech era.' Understanding the IPO process, its historical context, and why companies choose public vs private markets provides essential context for evaluating this strategic shift.
-
Colocation centre
1 min read
The article describes neoclouds as 'data center builders and resellers of AI chip clusters' - essentially a new form of colocation. Understanding the traditional colocation business model illuminates the skeptical VC critique that these are 'real-estate and power-market plays wrapped in an AI story'.
With neoclouds either already public (CoreWeave) or reportedly racing to do so (Lambda), these infrastructure-heavy AI cloud providers have become some of the biggest markups in the AI boom. And many top venture capital firms have largely missed out.
The rush to the public markets is unusual in the current tech era. Skeptical VCs snark that the neoclouds — essentially a new class of AI data center builders and resellers of AI chip clusters — are not really technology companies at all. Rather they’re real-estate and power-market plays wrapped in an AI story, with a boost from sweetheart relationships with Nvidia and other chipmakers.
Yet, as one AI-startup executive put it bluntly: “VCs are massively regretting not investing.”
Consider the action in the space just over past week weeks:
This excerpt is provided for preview purposes. Full article content is available on the original publication.