AnswersPublished17 sources
Why Jensen Huang Sees No Memory Chip Glut by 2029
In a CNBC interview on October 22, 2026, Jensen Huang said he saw no sign of a memory glut by 2029: AI systems can keep using more. Huang said HBM will remain in use for years and Nvidia will continue using LPDDR in data centers, citing its energy efficiency.
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer
Key takeaways
- In a CNBC interview on October 22, 2026, Jensen Huang said he saw no sign of a memory glut by 2029: AI systems can keep using more.
- Huang said HBM will remain in use for years and Nvidia will continue using LPDDR in data centers, citing its energy efficiency.
- Investor concerns center partly on Chinese supplier CXMT’s denser DRAM design and planned server memory product; Goldman Sachs argued the market may be overestimating how quickly China can close the technology gap.
AI Prompt
openai.comCreate a landscape editorial hero image for this Studio Global article: What did Nvidia CEO Jensen Huang say in his October 22, 2026, CNBC interview at GTC Berlin about why he sees no memory-chip oversupply by 20. Article summary: Huang rejected the prospect of a memory-chip glut by 2029, arguing that expanding AI use would keep absorbing more memory: “You can always use more.” [15] The available evidence supports that broad demand argument and co. Topic tags: general, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fa
Nvidia CEO Jensen Huang pushed back against forecasts of a memory-chip glut by 2029 in a CNBC interview at GTC Berlin on October 22, 2026. His argument was that demand can keep growing as AI systems use more memory: “You can always use more.” 6 That is a forecast, not proof that oversupply cannot happen—and concerns about new Chinese competition were already weighing on memory stocks.
6
32
Why Huang expects memory demand to keep growing
Huang’s central point was that memory remains useful as AI workloads expand. He said Nvidia would continue using large amounts of memory and argued that high-bandwidth memory (HBM) would remain relevant for years, rather than becoming obsolete in the near term. 6
23
He also pointed to Nvidia’s use of LPDDR, a type of memory associated with smartphones, in data centers. A report summarizing his remarks says he highlighted its energy efficiency, claiming it uses half the energy per unit of work compared with other options. 5 The available interview summaries support Nvidia’s continued use of LPDDR, but do not provide enough detail to independently assess the comparison or establish the full history of Nvidia’s adoption.
Why investors fear a glut
The near-term debate is not only about how much memory AI will need. Investors are also assessing whether new capacity and competition could eventually outpace demand. Memory stocks sold off in the week before Huang’s interview amid concerns that demand could weaken in the coming years. 6
One concern is ChangXin Memory Technologies (CXMT), a Chinese memory maker. Reports say the company made progress on a 4F² DRAM design, which reduces memory-cell area and can increase density, and was targeting a server-memory launch by the end of 2026. 35
33 A planned launch does not, by itself, establish production volume or prove that CXMT can match established suppliers at scale.
35
Those developments contributed to worries about greater competition for established memory companies, including Micron, SK Hynix and Samsung Electronics. 35
38 Goldman Sachs offered a more cautious counterpoint, arguing that investors had overestimated how quickly China could close the technology gap and describing two memory stocks as attractive on a risk/reward basis after the selloff.
32
A demand forecast, not a settled outcome
Huang’s view and investors’ concerns address different parts of the outlook. He expects AI’s appetite for memory—including HBM and LPDDR—to remain strong; investors are weighing whether weaker future demand or added supply could change the balance. 6
5
23
The evidence supports describing Huang’s position as confidence in sustained memory demand, not certainty about the market through 2029. CXMT’s reported design progress and planned product launch are reasons investors are watching supply competition, but the available reporting does not establish that those developments will produce an oversupply. 33
35
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "Why Jensen Huang Sees No Memory Chip Glut by 2029"?
In a CNBC interview on October 22, 2026, Jensen Huang said he saw no sign of a memory glut by 2029: AI systems can keep using more.
What are the key points to validate first?
In a CNBC interview on October 22, 2026, Jensen Huang said he saw no sign of a memory glut by 2029: AI systems can keep using more. Huang said HBM will remain in use for years and Nvidia will continue using LPDDR in data centers, citing its energy efficiency.
What should I do next in practice?
Investor concerns center partly on Chinese supplier CXMT’s denser DRAM design and planned server memory product; Goldman Sachs argued the market may be overestimating how quickly China can close the technology gap.
Sources
cnbc.comNvidia's Jensen Huang says he sees no sign of a 2029 memory glut: 'You can always use more'
finance.yahoo.comGoldman flags 2 memory stocks to buy amid deeper sector ...
qz.comJensen Huang pushed back on 2029 memory glut fears amid stock selloff
briefs.coNvidia: Memory Glut Unlikely by 2029, CEO Says