Surging AI Demand Tightens Storage Chip Supply

The rapid expansion of generative AI and large models means data centers need not only GPUs but also higher-capacity storage for training data, model parameters and inference results. Solidigm, SK hynix’s enterprise SSD business, said the volume of data generated by AI systems is growing faster than existing manufacturing capacity. The resulting supply-demand imbalance could become a bottleneck for infrastructure expansion.

Solidigm now expects tight storage chip supplies could persist through 2030, mainly because AI is driving a sharp increase in storage capacity requirements. NVIDIA’s chief executive has also said storage systems are under immense pressure, indicating that the AI server investment boom will also boost SSD demand and intensify scrutiny of suppliers’ capacity expansions and customers’ inventory-building.