If you were hoping memory chips would get cheaper anytime soon, SK Hynix has some bad news. CEO Kwak Noh-Jung is warning that the global memory chip shortage will persist beyond 2030, extending a supply crunch that’s already reshaping how the world’s biggest tech companies plan their infrastructure spending.
The timeline keeps getting longer. Back in March, SK Group Chairman Chey Tae-won told attendees at Nvidia’s GTC conference that wafer shortages could last until 2030, with supply trailing demand by more than 20%. Now the company is pushing that horizon even further out, suggesting the structural deficit in memory production is more entrenched than anyone initially estimated.
Why chips keep getting harder to find
The explosion in AI workloads has created insatiable demand for high-bandwidth memory, the specialized chips that power the GPUs and accelerators inside data centers.
New wafer capacity takes four to five years to come online. SK Hynix plans to double its memory wafer capacity over the next five years. The company announced that commitment in early June when Chairman Chey reiterated the shortage forecast.
