When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
Nvidia is urging SK Hynix to fast-track the production of its high-bandwidth memory (HBM4) chips as demand for AI hardware continues to soar.
According to a Bloomberg report, Nvidia CEO Jensen Huang made the request during a recent meeting with SK Group Chairman Chey Tae-won at the SK AI Summit in Seoul on November 4, 2024.
SK Hynix, a leading player in the global HBM chip market, had previously announced plans to supply its HBM4 chips to customers in the second half of 2025.
This timeline was already ahead of the initial target, but Nvidia’s request has pushed the delivery date forward by six months.
The urgency of Huang’s request reflects Nvidia’s dominant position in the AI chip market, where it currently holds about 80% of the market share ahead of the likes of Broadcom, Qualcomm and Intel.
The company’s GPUs are crucial for processing vast amounts of data required to train AI models, and the next iterations of these GPUs will incorporate the new HBM4 chips.
SK Hynix has been at the forefront of meeting the explosive demand for HBM chips, which are essential for AI technology. However, the company faces strong competition from other major memory manufacturers like Samsung and Micron, who are also investing heavily in high-bandwidth memory technology.
Samsung, for instance, plans to begin mass production of its HBM4 chips in the second half of 2025, after it tied a HBM4 AI chip development partnership deal with chip maker TMSC in September.
This move was largely perceived as a response to SK Hynix’s June announcement of a $75 billion investment in AI chips through 2028, with over half dedicated to high-bandwidth memory chips.
In addition to HBM4 chips, the South Korean memory chip maker is reportedly expanding its product lineup, with plans to supply a 12-layer HBM3E chip later this year and a 16-layer version in early 2025.
