SK hynix said on Wednesday that driven by surging artificial intelligence (AI) inference demand, it expects the performance scale of AI infrastructure to grow about 20-fold by 2030.
Ahn Hyun, head of SK hynix's development division, made the forecast during a keynote speech at the 2026 Semiconductor Exhibition (SEDEX), which opened earlier that day in southern Seoul and runs for three days.
Ahn said the AI inference market is expanding simultaneously in three areas: task complexity, the amount of information processed per request, and the number of inference requests. He said the interaction of these three factors is causing demand for AI infrastructure to grow rapidly.
As AI inference demand surges, memory is becoming increasingly important for storing and transmitting massive amounts of data. In particular, key-value (KV) caching—used to store and reuse information previously processed by AI models—is increasingly demanding capacity, requiring data to be distributed across different types of storage media, including high bandwidth memory (HBM), dynamic random access memory (DRAM), and solid-state drives (SSD).
The executive added that SK hynix is working to develop next-generation memory technologies including 3D stacked DRAM and high bandwidth flash (HBF), while continuing to improve the performance of its existing HBM products.
Comments