SK, Nvidia chiefs meet again as AI memory race intensifies

SK Group Chairman Chey Tae-won and Nvidia CEO Jensen Huang held their third meeting this year in Taipei, Taiwan, on Monday to strengthen cooperation on artificial intelligence memory. The discussions took place during Nvidia’s GTC Taipei 2026 conference and focused on SK hynix’s role as a critical supplier of high bandwidth memory (HBM), a component vital to AI infrastructure. The meeting follows earlier talks in Silicon Valley in February and San Jose in March, signaling a shift from a traditional supplier-customer relationship to broader collaboration on AI advancements. SK hynix shared photos of the gathering, which included senior executives from both companies, including SK hynix CEO Kwak Noh-jung. Nvidia’s upcoming Vera Rubin AI platform, set for release in the second half of 2026, will rely on sixth-generation HBM (HBM4), driving speculation that the company has significantly increased contract prices for memory supplies. Reports suggest potential price hikes of up to 100 percent, reflecting the growing demand for AI memory solutions. During the meeting, Huang emphasized performance, quality, reliability, and supply capacity as key priorities for HBM partnerships. Kwak Noh-jung downplayed speculation of immediate business negotiations, stating the discussions centered on long-term AI potential and strategic collaboration. Huang is scheduled to visit South Korea later this week for further meetings with Korean tech partners. The event also included a dinner with major Korean corporations, attended by executives from SK hynix, Samsung Electronics, LG Sciencepark, and Naver Cloud. SK hynix highlighted its recent achievement of a $1 trillion market capitalization, framing the partnership as pivotal to shaping the future of AI infrastructure.