AI-related equities continued a sharp sell-off in late June and early July 2026 amid growing skepticism regarding long-term enterprise returns [1, 3].
This trend signals a critical pivot in the AI boom. While chip demand remains high, the physical limitations of the U.S. power grid now threaten to bottleneck the expansion of the data centers required to run these systems [3].
Market volatility has hit major hardware providers. Nvidia stock fell about 18% from its 52-week high [4], with prices hovering around $190 [4]. Other chip makers, including AMD and Broadcom, have also faced pressure as investors question when massive AI investments will yield consistent profits [1].
Despite the chip slump, companies providing power infrastructure are seeing record growth. Eaton reported a 48% jump in its electrical backlog [2]. Similarly, Caterpillar saw its Power Generation revenue rise 41% year-over-year [2]. GE Vernova reported that data-center equipment orders hit $2.4 billion in the first quarter [2].
These gains come as the U.S. electricity grid struggles to keep pace with workload demands. Data-center power demand is projected to more than double between 2026 and 2030 [3]. Specifically, capacity is forecast to grow from 62,242 MW in March 2026 to 151,734 MW by March 2030 [3].
Grid operators now face the reality that projected demand may outpace available capacity [3]. This gap creates a risk where data centers cannot scale regardless of how many chips are available, potentially capping the growth of AI services across the country.
“Nvidia stock fell about 18% from its 52-week high”
The AI industry is moving from a phase of pure computational competition to a struggle over physical infrastructure. While the initial surge was driven by the availability of chips, the next ceiling is the electrical grid. If the U.S. cannot modernize power delivery at the speed of AI adoption, the resulting energy shortage could create a hard cap on the scaling of large language models and enterprise AI deployment.
