Episode 17: Cooling is the hard limit of AI

Share this article

As chips push past 2,000 watts, heat is emerging as the biggest threat to the future of AI. In this episode, Shahar Belkin, Chief Evangelist, at ZutaCore, explains why traditional cooling is failing, how two-phase solutions like pool boiling could redefine data centres, and why ignoring heat might stall the entire AI revolution.

Related Posts
Others have also viewed

The next frontier of start-up acceleration lies in the AI tech stack

The rise of generative and agentic AI has redefined what it means to start a ...

Quantum-centric supercomputing will redefine the AI stack

Executives building for the next decade face an awkward truth, the biggest AI breakthrough may ...

The invisible barrier that could decide the future of artificial intelligence

As AI workloads grow denser and data centres reach physical limits, the real bottleneck in ...
Into The madverse podcast

Episode 21: The Ethics Engine Inside AI

Philosopher-turned-AI leader Filippo explores why building AI that can work is not the same as ...