Jensen Huang, co-founder, president and CEO of US technology company Nvidia, gestures during a press conference on October 13, 2026 in Paris.
Joel Saget | AFP | Getty Images
Nvidia CEO Jensen Huang told CNBC on Thursday that it is unlikely there will be a glut in memory supply by 2029, even though memory stocks have sold off over the past week on concerns about oversupply and declining demand.
“One of the really great things about memory is that you can always use more memory,” Huang told CNBC’s Arjun Karpal in a wide-ranging interview on the sidelines of the company’s GTC Berlin conference. He added that HBM, a type of random access memory (RAM), will be used for many years to come.
“We were the first company in the world to introduce LPDDR memory (the kind found in mobile phones) into the world’s data centers, allowing us to deliver performance that is twice as energy efficient as any other memory in the world today. So we will continue to use large amounts of LPDDR memory as well,” he added.
Memory stocks have plunged over the past week after an op-ed column published in a prominent South Korean publication on Oct. 15 urged the country’s government to prepare for a slowdown in the AI boom.
The column quotes two experts who say they expect a “flood of new supply” to turn the memory market into a “red ocean” in 2029, as translated by CNBC. “The supercycle is over and the industry could enter a downturn by 2029,” they write.
Memory is a core component of Nvidia’s chips, and an oversupply could drive down prices and hurt profit margins for the companies that make the products. SK HynixSamsung, and micron.
space x and tesla CEO Elon Musk responded to the article at the time, saying that China “will be producing a lot of RAM in 2028.”
The decline in memory was also driven by China’s Changxin Memory Technologies (CXMT), which announced a new design architecture that could lead to improved efficiency. The announcement raised investor concerns that CXMT could steal market share from rivals SK Hynix, Samsung and Micron.
“We’re open-minded,” Huang said Thursday when asked if the company plans to partner with CXMT to use the company’s memory products. “We work with Chinese companies. Almost the entire world works with Chinese companies somewhere in the supply chain.”
“We need to take China’s capabilities very seriously because China is amazing in basic infrastructure technology and no country is better at overall manufacturing than China,” he added.
US regulations have ‘accelerated’ China’s chip progress
Fan also told CNBC that progress in China’s semiconductor industry is due to tighter U.S. regulation, forcing the country to speed up its chip technology and manufacturing.
The United States currently restricts exports of the country’s most advanced semiconductors, including top-of-the-line Nvidia chips, to China. But some Chinese companies are using loopholes to access the product’s computing power through data centers in Southeast Asia. Experts say this was the key to boosting the capabilities of China’s AI models.
China has “definitely accelerated because of all the restrictions,” he said.
“We have definitely accelerated China’s efforts, the nationalization of semiconductor manufacturing capacity, all the industrial support that has been put into that area, and there is no question that we have accelerated that,” he added.
Hwang stressed that American tech companies are “mighty” and ready to compete “with anyone.”
