LAS VEGAS — The term “AI factory” came up in multiple keynote presentations, media briefings, and in interviews with executives at this year’s . ”The reason I don’t call it a data center is that a data center is a place where people store data,” Huang said during a CES keynote.
“These AI factories are completely different from data centers,” Siemens CEO Roland Busch told reporters at the show. “They consume more energy, need liquid cooling, and cannot be controlled with normal building controllers — they need industrial-level controls.”
Booz Allen Hamilton CTO Bill Vass is on board with that particular definition. “An AI factory is a data center that is designed specifically for AI-specific hardware,” he explained. For example, it might have special concrete to support the physical weight of the server racks, and uses AI, digital twins and simulations to optimize power, cooling, and server placement.
Others, however, seem to use the word to mean something smaller than a data center, referring more to the servers, software, and other systems used to run AI.
For example, the . MIT Sloan Management Review’s Thomas H. Davenport and Randy Bean define it as a combination of “technology platforms, methods, data, and previously developed algorithms that make it fast and easy to build AI systems.”
“Not being able to build on an established foundation makes it both more expensive and more time-consuming to build AI at scale,” they wrote.
The .
SOCIAL SHARE CARD GENERATOR