Nvidia CEO Jensen Huang confirmed the new Vera data center CPU will use SK Hynix DRAM, expanding their partnership beyond HBM, and announced expanded deals expected to be "even larger" in the second half of the year a... SK Hynix is estimated to command 60–70% of HBM4 supply for Nvidia's Vera Rubin platform, and the...

Create a landscape editorial hero image for this Studio Global article: What did Nvidia CEO Jensen Huang confirm during his June 2026 Seoul visit about the Vera data-center CPU using SK Hynix DRAM, what broader p. Article summary: Here is a full breakdown of everything Jensen Huang confirmed during his June 2026 Seoul visit and the broader context.. Topic tags: general, general web. Reference image context from search candidates: Reference image 1: visual subject "Jensen Huang, CEO of NVIDIA, arrives at a Korean barbecue restaurant for a dinner meeting with SK Group Chairman Chey Tae-won, LG Group Chairman Koo Kwang-mo, and Naver Chairman Le" source context "Nvidia, SK to detail cooperation plan amid prolonged chip shortage" Reference image 2: visual subject "![Nvidia CEO Jensen Huang and SK Group Chairman Chey Tae-won, center, poses with executives from SK affiliates including SK hynix
Nvidia CEO Jensen Huang’s four-day tour of Seoul in early June 2026 wasn’t just a victory lap—it was a supply-chain summit disguised as a series of fried-chicken dinners. The visit produced concrete hardware confirmations, a freshly inked supplier trifecta for HBM4, and early signals of a partnership expansion that stretches from memory chips to humanoid robots.
The most concrete technical takeaway from the trip was Huang’s confirmation that Nvidia’s next-generation Vera data-center CPU will use DRAM chips from SK Hynix, pushing the companies’ relationship well beyond high-bandwidth memory . Huang told reporters outside a Seoul restaurant on Sunday, June 7, that Nvidia’s upcoming AI supercomputer lineups—Vera Rubin and the Vera CPU—would rely on SK Hynix DRAM
. The CPU is a central piece of the Vera Rubin architecture, which pairs a new Grace-family CPU with a next-generation GPU, and the choice of a Korean memory partner for the CPU socket signals strategic supply-chain alignment.
Huang used the visit to settle months of supply-chain speculation around HBM4. Stepping off his plane at Seoul’s Gimpo airport on June 5, he confirmed that all three major memory manufacturers—Samsung, SK Hynix, and Micron—have been certified and are in active production for Nvidia’s Vera Rubin HBM4 slots . The first systems are scheduled to ship in Q3 2026
.
Industry analysts estimate SK Hynix holds roughly 60–70% of the HBM4 volume allocated to Vera Rubin, with Samsung at approximately 25–30% and Micron supplying the remainder . Nvidia has not disclosed official allocation figures, and earlier in 2026 analyst firm SemiAnalysis had projected Micron’s share at 0%—a question the June 5 certification of all three suppliers appeared to settle definitively
.
Huang’s Seoul schedule culminated in a Monday-morning tea meeting on June 8 with SK Group Chairman Chey Tae-won at the SK Seorin Building in Jongno, their fifth gathering of the week . SK Hynix confirmed that Chey and Huang planned to brief the media together afterward
. The expected collaboration plan covers procurement and supply of high-performance memory for AI, with Huang signaling that the deals will be “even larger in the second half and next year”
.
The partnership blueprint likely includes:
Huang framed the alliance expansively, stating Nvidia is “growing together with Hyundai Motor, LG, SK Hynix, Samsung, and Naver” across the full AI stack . He brought what he called a “big gift” to Korea: four new businesses—Vera Rubin, Vera, RTX Spark, and Jetson Thor—along with plans for a prominent Nvidia research center in the country
.
Huang declared robotics as South Korea’s “next major sector” and promised “surprises” for Korea . He highlighted that Korea is an “indispensable core axis” of Nvidia’s AI ecosystem and expressed intent to expand collaboration beyond semiconductors into robotics and AI factories
. In a notable aside, Huang suggested Seoul could host Nvidia’s annual GTC developer conference in the future—a signal of how central the Korean market has become to Nvidia’s physical AI ambitions
.
Away from the fried-chicken diplomacy, Huang issued a stark warning: the memory shortage will last “quite a few years” as AI demand far outpaces supply . SK Hynix’s HBM capacity is already sold out through 2027, and some customers are offering advance payments of 10–30% to secure capacity
. Against this backdrop, the company’s financials are remarkable:
To fund factory expansion in Yong-in, M15X, and a packaging facility in Indiana, SK Hynix has confidentially filed a Form F-1 with the SEC for a U.S. ADR listing targeting the second half of 2026, expected to raise an estimated $10 billion to $14 billion . Chairman Chey first confirmed the listing review at GTC 2026 in March, calling it a move to make SK Hynix “a more global company,” and the company has since told investors it received “tremendously positive” feedback on the plan
.
With the Vera Rubin platform in full production, the Vera CPU locked to SK Hynix DRAM, and a U.S. listing on the horizon, Huang’s Seoul trip marked the moment Nvidia’s AI supply chain went from transactional to structural—anchored in Korean manufacturing and capital markets.
Studio Global AI
Use this topic as a starting point for a fresh source-backed answer, then compare citations before you share it.
Nvidia CEO Jensen Huang confirmed the new Vera data center CPU will use SK Hynix DRAM, expanding their partnership beyond HBM, and announced expanded deals expected to be "even larger" in the second half of the year a...
Nvidia CEO Jensen Huang confirmed the new Vera data center CPU will use SK Hynix DRAM, expanding their partnership beyond HBM, and announced expanded deals expected to be "even larger" in the second half of the year a... SK Hynix is estimated to command 60–70% of HBM4 supply for Nvidia's Vera Rubin platform, and the company is preparing a U.S.
Huang declared robotics South Korea’s "next major sector" and teased that Seoul could host Nvidia’s GTC conference in the future, signaling a broad alliance stretching from AI supercomputers to digital twins.