
"The current shortage could continue until 2030. AI actually wants to have a lot of HBM, and once you make the HBM, we have to use a lot of wafers. The industry faces a wafer deficit of more than 20% and at least four to five years of capacity building lie ahead before supply can match demand."
"This is no longer a cyclical imbalance. It is a structural reallocation of the memory market driven by AI infrastructure economics. The biggest mistake right now is to view this as a wafer or DRAM shortage. The constraint is systemic."
The global semiconductor industry faces a prolonged memory shortage extending to 2030, driven primarily by artificial intelligence infrastructure demands. SK Group Chairman Chey Tae-won reports a wafer deficit exceeding 20%, requiring four to five years of capacity expansion before supply meets demand. AI systems require substantial high-bandwidth memory, consuming significant wafer resources. SK Hynix controls 57% of the global HBM market and 32% of DRAM supply. Industry analysts characterize this as a structural market reallocation rather than a cyclical shortage, with systemic constraints affecting memory pricing and availability. SK Hynix plans to announce strategies for stabilizing DRAM prices, though details remain undisclosed.
#semiconductor-shortage #ai-infrastructure-demand #memory-supply-constraints #dram-pricing #hbm-market
Read at Computerworld
Unable to calculate read time
Collection
[
|
...
]