What Happens When AI Leaders Need More Memory?

🔍 Read the full analysis: What Happens When AI Leaders Need More Memory? on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

• Business-only prices and quantity discounts
• Tax-exempt purchasing
• Multiple users, one account, clear invoices

Create a free Business account →Free to register

As an affiliate, we earn on qualifying purchases.Hide

OpenAI CEO Sam Altman met about 50 senior Samsung executives in Yongin, South Korea, on Oct. 31, as AI companies compete for memory chips used in data centers. Samsung said the meeting was part of a leadership program and discussed AI’s business impact; neither company announced a new memory-supply agreement.

Samsung said participants came from its electronics, finance and biotech affiliates and met Altman at the company’s Human Resources Development Center, amid growing attention to AI’s memory needs. The stated theme was “The Next 10 Years Shaped by AI: The Enterprise’s Choice.” Samsung’s account focused on AI’s direction, its effect on management decisions and potential new markets. Yonhap, citing Samsung, reported that Altman told the group that using AI to move faster and improve products matters more than simply adopting it.

Separate reporting tied the visit to the supply race. Bloomberg said the meeting aimed to deepen cooperation on memory chips and AI infrastructure as shortages intensify. The Seoul Economic Daily reported that OpenAI executive Richard Ho, who leads its in-house chip work, accompanied Altman, and that OpenAI’s planned accelerator may use Samsung’s next-generation HBM4 memory. Samsung and OpenAI have not confirmed the HBM4 report, and no new deal, price or supply volume was disclosed.

The visit followed another meeting in Seoul on Oct. 31: SK Group chairman Chey Tae-won met Altman alongside executives from SK hynix, according to Yonhap, which cited industry sources. The parties did not disclose what they discussed. SK hynix, like Samsung, is a major memory-chip supplier.

AI Demand Meets Memory Limits

Memory availability affects how quickly companies can build and expand AI data centers, where large quantities of DRAM and high-bandwidth memory support computing systems. When supply is tight, chipmakers’ capacity and allocation decisions can shape infrastructure costs and the pace of new deployments. Altman’s meeting is relevant because OpenAI is seeking the hardware needed to run and develop AI systems, while Samsung is a leading memory producer.

The consequences may also reach buyers outside the data-center market. TrendForce forecasts that conventional DRAM contract prices will rise 10% to 15% in the fourth quarter of 2026. It also warns that PC memory supply could decline in 2027 as more capacity shifts toward servers. Those forecasts do not establish what any particular device will cost, but they point to possible pressure on manufacturers and consumers through higher prices or changes to product specifications.

Altman’s argument about PC memory adds an allocation debate to the supply story. If manufacturers direct more output to large AI customers, people buying computers could face tighter supply or higher costs. That outcome is not established by the meeting, and Samsung has not announced that it will reduce consumer supply. The key issue is how limited production is divided between data centers and devices.

Products Worth Considering

DJI Osmo Mobile 8 Gimbal with Mic Mini Transmitter, Infinity Black

AmazonView Latest Price

Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite

AmazonView Latest Price

Apple Watch Ultra 4 GPS + Cellular 49mm Smartwatch

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

A Partnership Built Before the Visit

The meeting took place against an existing business relationship. In October 2025, OpenAI and four Samsung companies signed a letter of intent. Samsung said at the time that OpenAI’s projected memory demand could reach up to 900,000 DRAM wafers per month. That was a projection in a letter of intent, not a confirmed monthly purchase commitment or a newly announced supply contract.

In June 2026, OpenAI said Samsung Electronics would introduce ChatGPT Enterprise and Codex for employees in Korea and its global device division. The Korea Herald reported that Samsung executive chairman Lee Jae-yong visited OpenAI in San Francisco in July. Altman’s Oct. 31 visit was his first to Samsung in about a year, according to the Seoul Economic Daily.

The memory market is also becoming more central to Samsung’s results. In the company’s third quarter, its chip division recorded 106.5 trillion won in operating profit, out of 107.4 trillion won companywide, according to the source material. Samsung’s phone and network business reported an operating loss of 0.6 trillion won, partly attributed to cost pressures. Those figures show the weight of chips in the company’s recent results; they do not, by themselves, establish the cause of any future supply decision.

“From a global allocation perspective, it’s quite terrible that so much of the world’s limited memory sits in people’s own computers, being used 5% of the time.”

— Sam Altman, on the Cheeky Pint podcast, published Oct. 29

Supply Terms Remain Undisclosed

There is no public confirmation of a new OpenAI-Samsung supply contract, the amount of memory either company discussed, or any prices or delivery schedule. Samsung’s account described the event as leadership training, while Bloomberg reported a focus on deeper cooperation in memory and AI infrastructure. The accounts differ in emphasis, and the meeting’s full agenda and attendee list have not been released.

The reported use of Samsung HBM4 in an OpenAI accelerator remains unconfirmed by both companies. It is also unclear how much of the 2025 projected demand will become purchases, how much memory Samsung will allocate to data centers rather than PCs and other devices, and whether forecast supply constraints will lead to specific price increases for consumers. Altman’s estimate that PC memory is used 5% of the time was not presented as an independently measured figure.

Watch for Deals and Pricing

The next concrete indicators will be any company announcements specifying memory supply, chip technology, delivery volumes or commercial terms. Until such details emerge, the Oct. 31 meeting is evidence of continued engagement, not proof that OpenAI has secured additional capacity.

Buyers and manufacturers will also be watching subsequent memory-market updates, including whether TrendForce’s fourth-quarter DRAM price forecast and warning about 2027 PC supply are borne out. Future results from Samsung and other chipmakers may offer more information about demand and capacity, but the source material provides no announced date for a follow-up meeting or supply decision.

Key Questions

What was confirmed about Altman’s meeting with Samsung?

Samsung said about 50 executives from its electronics, finance and biotech affiliates met Altman in Yongin on Oct. 31. It described the event as part of a leadership program focused on AI.

Did OpenAI and Samsung announce a new memory deal?

No. No new contract, supply volume, price or delivery schedule was announced. The companies had previously signed a letter of intent in October 2025.

What did Altman say about memory in personal computers?

On the Cheeky Pint podcast, Altman argued that memory in personal computers is idle much of the time and could be used more fully in shared data centers. His 5% figure was an estimate, not an independently reported measurement.

Could the memory shortage affect PC prices or specifications?

It could, but a specific effect has not been confirmed. TrendForce forecasts conventional DRAM contract prices will rise 10% to 15% in the fourth quarter of 2026 and warns PC memory supply could decline in 2027 as capacity shifts toward servers.

Source: ThorstenMeyerAI.com