Databricks’ latest fundraising efforts hinge on surging demand for artificial intelligence infrastructure and a global shortage of the computing power needed to run AI models, according to CEO Ali Ghodsi.

Speaking on CNBC’s "The Exchange," Ghodsi said the latest funding round will help Databricks expand its AI infrastructure capabilities and keep pace with growing demand from customers across regions including the U.S., Asia and India.

"We are running out of GPUs everywhere," Ghodsi said. "We ran out of GPU almost in Asia. Every country in Japan, Korea, U.S., India — so we just needed to go get a lot of these, and that requires a lot of funding."

The data analytics and AI software company currently boasts a $188 billion valuation.

Returns Don’t Match Pace Of Spending

The comments highlight a growing constraint across the AI industry: securing enough advanced graphics processing units (GPUs), the critical chips required to train and run large AI models. As companies move beyond AI experimentation and begin deploying the technology across their businesses, demand for computing capacity has surged.

Ghodsi also highlighted growing concerns among businesses about the economics of AI adoption. While companies are increasing investments in AI technology, some are struggling to see returns that match the pace of their spending.

"One executive of a Fortune 500 told me his cost of AI is increasing exponentially, [but] his revenue is not increasing," Ghodsi said. "If this continues, he will go bankrupt."

That pressure is pushing companies toward a more cost-conscious approach to AI, where organizations use expensive frontier models for complex tasks while turning to lower-cost alternatives for more routine workloads.

Ghodsi said open-source models, including China-based Kimi, are gaining traction because they allow companies to expand AI usage while managing costs.

"For the really difficult questions, let’s ask the frontier American models and pay a lot for it," Ghodsi said. "But for most of these mundane tasks… use a cheap source."

Photo Courtesy: bluestork on Shutterstock.com