Nvidia has secured a long-term supply of advanced artificial intelligence (AI) memory from South Korea's SK Hynix as part of a US$500 billion agreement aimed at expanding global AI infrastructure and easing concerns over shortages of critical components.
The deal, announced in San Francisco on Friday, includes the construction of large-scale AI data centres expected to begin operating in 2027.
Nvidia said the facilities will require around 2 gigawatts of power, underscoring the scale of the project, which is expected to deploy hundreds of thousands of graphics processing units (GPUs).
According to the press release, "Today’s announcement builds on the decades-long technology partnership between SK Group and NVIDIA, including the recently announced plans for SK Telecom to build a 2-gigawatt-scale AI Cloud in Korea.
"This cloud will use the NVIDIA® DSX™ platform and deploy NVIDIA Vera Rubin accelerated computing powered by SK hynix HBM4, with the first AI factory planned to come online in 2027.
"It will be built on the NVIDIA DSX full-stack AI factory architecture, which integrates NVIDIA accelerated computing, systems, software and partner technologies to deliver the lowest token cost at maximum energy efficiency.
“The two companies aim to accelerate large-scale AI infrastructure development, including sovereign, physical, agentic and enterprise AI services, and jointly address the increasing AI demand across the Asia-Pacific region, including South Korea.”
The agreement strengthens Nvidia's efforts to secure supplies of high-bandwidth memory (HBM), a critical component used in its AI chips.
Demand for HBM has surged alongside the rapid expansion of artificial intelligence, creating a global supply shortage. Industry analysts regard SK Hynix as the leading producer of HBM technology.
"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, was quoted as saying in this CNBC story.
Separately, Samsung Electronics announced it had signed a memorandum of understanding with Broadcom to expand collaboration across advanced memory and semiconductor foundry technologies.
Samsung said the partnership, estimated to be worth $200 billion, is designed to support the next generation of AI infrastructure by strengthening chip manufacturing and memory capabilities.
Nvidia also revealed plans to invest $1 billion in South Korean cloud provider Naver, which is developing AI data centres powered by Nvidia GPUs.
The company said the project will allow customers in South Korea and internationally to secure AI computing capacity before the first 200 megawatts of infrastructure comes online.
