Mirantis partners with neocloud providers to build AI factories

Mirantis said it partnered with a new class of Artificial Intelligence (AI) infrastructure providers called “neoclouds” to build and operate AI factories intended to produce intelligence at scale.

The announcement cited a Goldman Sachs forecast of $100 trillion in AI infrastructure investment by 2040 and reported that some observers questioned whether the planned buildout was sustainable.

The release described neoclouds as providers that secured energy access, built computational capacity, and sold intelligence services; it also included per-100-megawatt revenue examples: bare metal Graphics Processing Unit (GPU) infrastructure might generate $1 billion, orchestrated just-in-time compute $1.5 billion, and packaged AI services $4 billion.

Mirantis said it worked with emerging neocloud providers and noted those providers had acquired land, negotiated grid connections, and secured power contracts as part of their infrastructure development.

“Right now, we're seeing two primary applications driving all this infrastructure demand: consumer AI like ChatGPT, and enterprise productivity tools like coding assistants,” said Freedland. “Companies that control energy and know how to convert it efficiently into intelligence have real competitive advantages,” said Freedland. “This isn't just about data privacy, though that's part of it,” said Freedland.

He predicted hybrid consumption models would become standard, that enterprises would consume AI services from multiple sources including hyperscalers for commodity intelligence, neoclouds for specialized capabilities, and sovereign providers for strategic applications, and he said he questioned whether current projections were ambitious enough.