AI-related stocks have been volatile in recent weeks as investors debate whether the sector is facing a demand slowdown or simply a pause after a sharp rally. But several executives speaking to CNBC said they are not seeing signs of excess capacity in the broader buildout of AI infrastructure.
Pat Gelsinger, the former Intel chief executive and now a general partner at Playground Global, said he sees AI demand as “almost unlimited,” with energy availability the main constraint. Marc Boroditsky, chief revenue officer at Nebius, said demand for compute has been “extraordinary” and continues to exceed what the company can supply. Andrew Feldman, chief executive of Cerebras Systems, said cases such as Meta and xAI renting out surplus capacity appear to be isolated rather than evidence of a wider glut.
Other companies echoed that view. Sungyun Park, chief executive of Rebellions, said momentum in AI infrastructure remains strong and argued that the recent developments do not show hyperscalers are broadly overinvesting. Michael Hurlston, chief executive of Lumentum, said the company’s products are sold out for the next five years and that it is trying to expand capacity to meet demand.
At the same time, executives acknowledged that enterprises are becoming more focused on the economics of AI use. The report said businesses are moving from a period of heavy experimentation toward a more measured approach that emphasizes value and return on investment, even if overall interest in AI remains high.
Source: cnbc.com
