AI & the Cloud

With AI and the Cloud expanding at such a rapid pace, I looked up some information as to what is involved in AI and the Cloud. It’s an interesting name, the Cloud, but our stored information doesn’t reside upstairs in a cloud, but downstairs in mega-acre storage data centers

Leading companies are using up to 16,000 specialized chips and large-scale supercomputers packed with hundreds or thousands of specialized chips with 4-8 high-end GPUs (Graphic Processing Units) per server.

Major tech companies, eg, (META) power their private clouds with tens of thousands of NVIDIA H100 GPUs to perform trillions of operations per second.

This new need for power is creating new Data Centers averaging 1 million square feet, with some exceeding 10 million square feet.

There are 50,000-acre Data Centers in early-stage development, which can consume enough power for 5 million homes, and are facing challenging supply chain disruptions.

Cooling Data Centers can account for 40% of electrical demand and require a lot of water…which can be scarce.

Not only is Artificial Intelligence eliminating certain jobs, but it also necessitates an extraordinary need for electricity, water, and space. Can our infrastructure keep up with the demand?