Will China Banned Nvidia Chips Impact the Global RAM Market?

Tesla CEO Elon Musk sparked investor anxiety on October 15 by warning that China will be "making a lot" of random-access memory (RAM) by 2028, a prediction that suggests a looming global supply glut. This forecast directly challenges the current dominance of South Korean chipmakers and could reshape the semiconductor landscape as China scales its production capabilities.
- Samsung Electronics shares dropped 10.56 percent on October 15 following Musk's comments, according to koreajoongangdaily.com.
- SK hynix shares fell 12.76 percent on the same day, reported by koreajoongangdaily.com.
- Gartner projects global semiconductor revenue will reach $1.6 trillion in 2026, according to tradingview.com.
Is a global chip glut actually coming?
The semiconductor industry is currently split between those predicting a massive oversupply and those expecting endless demand driven by artificial intelligence. While Musk anticipates a glut as early as 2028, other industry leaders see a different future. The disagreement has caused significant volatility in the stock markets of major chip producers.
Nvidia CEO Jensen Huang argues that memory supplies are unlikely to exceed demand even by 2029. He believes the high bandwidth memory (HBM) required for AI processors will remain in heavy use for many years. This view suggests that the AI boom provides a massive cushion that could absorb any new production entering the market.
However, some analysts worry about the long-term stability of this boom. A column in South Korea's Maeil Business Newspaper warned that a wave of new supply could flood the market by 2029. This potential oversupply could depress prices and intensify competition, particularly for companies that do not specialize in high-end AI chips.
How does Elon Musk see the Chinese threat?
Elon Musk views China's entry into the memory market as a potential disruptor that could break the current oligopoly. He believes China's ability to scale production once they master a technology is unmatched. This could lead to a scenario where standard memory components become significantly cheaper, hurting the profit margins of established players.
Musk's warning followed a report suggesting South Korea's reliance on Samsung Electronics and SK hynix could be exposed by 2029. He accelerated this timeline by suggesting the Chinese expansion would be in full swing by 2028. His comments suggest a shift from a market controlled by three major players to one with much more competition.
Elon Musk cautioned that China will be "making a lot" of random-access memory (RAM) by 2028.
The scale of this potential shift is supported by data from Morgan Stanley. According to news.qq.com, China's ChangXin Memory Technologies (CXMT) could see its DRAM monthly wafer capacity grow from 180,000 in 2025 to 500,000 by 2028. This would position the company to potentially surpass Micron as a top global supplier.
Can China overcome US technology bans?
The primary obstacle for Chinese chipmakers is the lack of access to advanced manufacturing tools. Specifically, US regulations prevent Chinese firms from acquiring extreme ultraviolet (EUV) lithography equipment. This equipment is essential for making the most advanced and efficient chips in the world.
Because they cannot access EUV, Chinese companies like CXMT are forced to find alternative ways to compete. One method involves using "4F-squared" designs to shrink memory cells without the most advanced tools. This allows them to pack more data onto a chip of the same size, helping them stay competitive in the standard memory market.
Experts are divided on whether this can work. Roh Geun-chang, CEO of Semicon Research Lab, noted that the acquisition ban imposes a substantial barrier to technological competitiveness. He stated that even if Chinese companies announce expansion plans, the actual increase in high-end capacity will likely be limited because major players like Samsung and TSMC are buying up the available equipment.
On the other hand, CXMT's CEO Cao Kanyu stated that the company has developed a new transistor to address electrical leakage. This technology is intended to lay the groundwork for the 4F-squared design. The company aims to launch a DDR5 server memory module using this method by the end of this year.
What are the risks for South Korean manufacturers?
South Korean companies face a dual threat: the rise of Chinese competitors and the financial stability of their customers. While Samsung and SK hynix currently lead in high-bandwidth memory (HBM), they remain vulnerable in the high-volume, lower-margin standard memory markets where China is expanding rapidly.
The risk is not just about competition, but also about who is buying the chips. Smaller, debt-funded cloud providers like CoreWeave have less financial flexibility than giants like Google or Amazon. If AI demand falls or financing becomes more expensive, these smaller providers could struggle, leading to a shakeout that hits all chip suppliers.
Currently, demand remains high. Executives at Amazon and Google have stated their data center capacity is fully booked. Local media reported on Tuesday that Chinese cloud providers, including Alibaba and Tencent, are seeking long-term supply deals with Korean firms, sometimes asking for contracts to extend as far as 2035.
How does the market view the current boom?
Despite the fears of a glut, the financial outlook for the semiconductor industry remains largely positive in the short term. South Korean Finance Minister Lee Hyoung-Il told CNBC that the industry is on a strong growth path through at least 2027. He noted that demand continues to outstrip supply by a wide margin.
Market projections suggest massive growth is still ahead. Bank of America projects the global semiconductor market could nearly double to $3.2 trillion by 2030, according to tradingview.com. This growth is expected to be driven by memory chips and components for data centers.
The volatility in stock prices reflects the uncertainty of this timeline. While Samsung and SK hynix saw significant drops following Musk's remarks, they also saw a rebound after new filings from AI companies like Anthropic. This shows that investor sentiment is highly sensitive to both supply fears and AI demand news.
What happens next?
The semiconductor industry will closely monitor several key developments over the next two years. Investors will be looking for signs of whether China's production capacity for DDR5 and other server memory modules actually hits the targets set for 2025 and 2026. Any breakthrough in Chinese manufacturing efficiency could accelerate the timeline for a market glut.
Additionally, the industry will watch the capacity expansion of ASML. As the sole provider of EUV machines, ASML is expected to increase its production capacity from 44 units in 2025 to over 80 units in 2027. The ability of major manufacturers to secure this equipment will determine if the technological gap between them and Chinese firms remains wide or begins to close.
Finally, the continued capital expenditure of "hyperscalers" will be critical. If major AI players continue to borrow heavily to build data centers, the high demand may prevent a price crash regardless of how much Chinese supply enters the market.
Frequently Asked Questions
Why did Elon Musk warn about a chip glut?
Musk believes that China's rapid scaling of memory chip production will lead to a massive oversupply by 2028, which could crash prices and hurt current market leaders.
Can China compete with Samsung and SK hynix?
China is becoming a major competitor in standard memory (RAM), but it still struggles to compete in the advanced high-bandwidth memory (HBM) used for high-end AI processors.
How do US sanctions affect Chinese chip production?
US sanctions prevent Chinese companies from buying the advanced EUV machines needed to make the most sophisticated chips, forcing them to use older or less efficient manufacturing methods.


