AI executives say compute demand remains strong despite chip-stock volatility

According to CNBC, several AI executives said demand for AI computing remains strong despite renewed volatility in chip and AI data center-related stocks, while noting businesses are becoming more cautious about AI costs. Former Intel CEO Pat Gelsinger, now a general partner at Playground Global, said he views AI demand as nearly unlimited and that energy availability is the main constraint. Meta’s announcement that it will sell excess AI computing capacity and Elon Musk’s xAI renting out excess capacity contributed to a sell-off in some related names, while Samsung forecast a sharp rise in profit but its shares fell after a more than 360% rally over the past 12 months. Nebius Chief Revenue Officer Marc Boroditsky said demand is exceeding what the company can fulfill as it builds data centers using Nvidia GPUs, and Cerebras Systems CEO Andrew Feldman said industrywide compute demand far outstrips available capacity and that the Meta and xAI situations are unusual. Rebellions CEO Sungyun Park said AI infrastructure momentum remains strong and he does not see the capacity moves as a signal that hyperscalers are overinvesting. Lumentum CEO Michael Hurlston said the company’s photonics and optical products for data center connectivity are sold out for the next five years, and the stock is up about 600% over the last 12 months. Boroditsky also said enterprise AI spending is shifting from heavy usage toward more ROI-focused “rationalization,” as companies weigh expensive frontier models against open-source alternatives such as DeepSeek or Alibaba.