Nvidia CEO highlights memory constraints in AI, boosting Micron and Sandisk shares amid rising data center capacity needs.
Nvidia CEO Jensen Huang identified memory as the primary bottleneck for AI advancement during CES 2026, citing growing demands for data center capacity. Large language models now require expanded memory to retain user interactions and context, straining existing infrastructure.
Since Huang’s remarks, Micron Technology (MU) and Sandisk (SNDK) have outperformed Nvidia in stock performance. Both companies reported surging revenue tied to AI memory solutions, reflecting heightened demand for high-bandwidth memory (HBM) and storage.
The shift underscores a broader industry pivot toward addressing memory constraints as AI adoption accelerates, with data centers prioritizing scalable solutions to support long-term model retention.