Seoul:SK hynix Inc. has projected that the performance scale of artificial intelligence (AI) infrastructure will expand approximately 20-fold by the year 2030, driven by the rapidly increasing demand for AI inference.
According to Yonhap News Agency, Ahn Hyun, the head of development at the South Korean chip giant, shared this projection during his keynote speech at the 2026 Semiconductor Exhibition (SEDEX), which commenced earlier in the day for a three-day event in southern Seoul.
Ahn highlighted that the AI inference market is expanding across three main areas task complexity, the amount of information processed simultaneously, and the volume of inference requests. These factors are collectively driving the demand for AI infrastructure at a rapid pace.
As the demand for AI inference continues to surge, the importance of memory for storing and transferring large volumes of data is increasing. Particularly, the growing capacity needs of key-value (KV) caches, which store and reuse information processed by AI models, necessitate data distribution across various types of memory, such as high-bandwidth memory (HBM), dynamic random-access memory (DRAM), and solid-state drives (SSDs).
SK hynix is actively pursuing advancements in next-generation memory technologies, including 3D stacked DRAM and high-bandwidth flash (HBF), while also working to enhance the performance of existing HBM products, Ahn added.