SK Hynix to double wafer capacity over five years

SK Hynix plans to double its wafer production capacity over the next five years to meet rising demand from artificial intelligence and data-intensive computing, SK Group Chairman Chey Tae-won announced Tuesday at the Computex conference in Taipei.

He framed the capacity expansion as a main objective as the company builds manufacturing capability to support future product complexity and higher volumes.

Chey warned that a global wafer shortage could persist until 2030 and urged broader partnerships within Taiwan’s semiconductor ecosystem, saying SK Hynix should not rely only on cooperation with TSMC.

He highlighted the company’s role in supplying high-bandwidth memory for AI systems and expressed hope SK Hynix ‘can become a major supplier of high-bandwidth memory (HBM) for Nvidia’s Vera Rubin system.’ HBM is advanced memory used in AI accelerators and high-performance computing.

Industry data show SK Hynix held about 58% of the global HBM market in the first quarter, while Samsung Electronics and Micron Technology each had roughly 21%.

Last week SK Hynix surpassed a market value of $1 trillion for the first time, joining Samsung and Micron. Semiconductor stocks have risen amid expectations that AI will increase long-term demand for advanced memory products.

Goldman Sachs raised its 2028 operating profit forecast for SK Hynix by 24% to 454 trillion won (about $299.6 billion) and increased its forecast for Samsung Electronics by 23.3% to 610 trillion won.

SK Hynix is the memory-chip unit of South Korea’s SK Group. The Computex remarks follow months of industry discussion about supply shortages and shifts in the memory market as AI applications drive demand for specific chip categories.