South Korea's president flew to San Francisco to seal nearly $1 trillion in AI agreements. Nvidia locked down memory supply, cloud partnerships, and autonomous vehicle deals.
South Korean President Lee Jae Myung flew to San Francisco on July 24 for a summit with the most powerful names in artificial intelligence. Jensen Huang was there. Sam Altman was there. The heads of Samsung, SK Group, Hyundai Motor and Naver joined them. By the end of the day, roughly $950 billion in new AI agreements had been signed. South Korea positioned itself as the country most central to the next phase of the AI buildout.
Nvidia (NVDA) locked down AI memory supply from SK Hynix, South Korea's second most valuable company, CNBC reported. The agreement, unveiled late that evening in San Francisco, could be worth $500 billion over a number of years. It includes large-scale data centers expected online in 2027. SK Hynix affiliate SK Telecom will build a cloud business using Nvidia's Vera Rubin systems as part of the deal.
Nvidia said it is targeting enough capacity to require 2 gigawatts of power. That figure points to a buildout involving hundreds of thousands of graphics processing units working together.
"The expansion will include a co-develop opportunity for us on the next-generation SK Hynix AI memory, and this will help us secure a stable supply of HBM memory," Raj Mirpuri, Nvidia's enterprise vice president, told reporters on a call. High bandwidth memory, known as HBM, sits directly next to AI chips and feeds them data fast enough to keep expensive processors from sitting idle.
SK Hynix has built its recent momentum on exactly that product. The company ranked first globally in HBM revenue with a 56.4% share in the first quarter of 2026. Its Nasdaq debut showed first-quarter revenue reaching 52.58 trillion won, roughly $34.5 billion, up 198% from a year earlier, TheStreet reported. The stock now trades under the ticker SKHY.
A second, unrelated deal landed the same day. Samsung Electronics signed a memorandum of understanding with chip designer Broadcom to expand their collaboration across memory and foundry technologies. The agreement is worth an estimated $200 billion, Reuters reported. It is meant to help support the next generation of AI infrastructure.
Broadcom already builds custom silicon for Google's TPU program. Adding Samsung's memory and foundry capacity gives Broadcom another lever to pull. For Samsung, the deal is part of a broader push to close the gap with SK Hynix in HBM production while rebuilding its contract manufacturing business. Samsung has also been courting AI labs directly, a pattern where memory makers move beyond supplying parts into designing the systems that use them.
Nvidia's South Korea push was not limited to memory chips. The company said on July 24 it would invest $1 billion into Naver, a Korean cloud company building data centers around Nvidia's GPUs. The project is intended to triple the facility size and provide 200 megawatts of AI computing capacity.
Nvidia chief executive Jensen Huang told the gathering that the SK Group partnerships alone represented more than $500 billion in combined business. He did not detail how that figure was calculated, according to the Korea Herald. "The SK Group and I are announcing today that our two companies will enter into business partnerships that will represent over $500 billion of business together," Huang said during a meeting with President Lee in San Francisco.
Huang also said Nvidia would work with Hyundai Motor Group on autonomous vehicles and robotic systems. That extended the day's announcements beyond data centers and semiconductors into transportation.
The deals involve foreign governments and industrial conglomerates writing checks at a scale similar to American hyperscalers.
The memory market that underpins all of these deals remains a three-way race. Samsung Electronics leads with a 38% share of the broader DRAM market. SK Hynix holds 29%. Micron has 22%, according to Counterpoint Research data cited by TheStreet. Every one of those three now has direct exposure to Nvidia's expanding supply chain.
Analysts have been warning for months that memory demand is outpacing supply. The announcements on July 24 only add to that pressure. BofA has pointed to the same dynamic as a reason memory prices could climb across DRAM, NAND and HBM products alike.
At AlphaScala, NVDA holds a score of 75 out of 100, classified as "Moderate," with the stock trading at $206.84, down 0.92% on the session.
For U.S. investors without direct access to Samsung shares, SK Hynix's new Nasdaq listing and Micron remain the two clearest ways to track the trade. Both are now tied more tightly than ever to decisions being made inside Nvidia's San Francisco headquarters.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.