Samsung, SK Hynix Unveil $950 Billion US Chip Supply Pact

by priyanka.patel tech editor

South Korea’s Samsung Electronics and SK Hynix announced $950 billion in U.S. chip supply partnerships during an AI summit in San Francisco, with CNBC deals and SK Group pledging $750 billion in AI infrastructure investments.

South Korea’s semiconductor giants Samsung Electronics and SK Hynix unveiled a $950 billion pact with U.S. technology firms during an AI summit in San Francisco, marking a pivotal shift in global chip supply dynamics. The agreements, announced July 24, 2026, involve long-term memory chip supplies, data center construction, and AI infrastructure development, with Nvidia securing a $500 billion deal to lock in high-bandwidth memory (HBM) from SK Hynix. The collaborations aim to address a surging demand for AI computing power, as companies race to outpace a global memory shortage.

Nvidia’s $500 Billion HBM Pact and Data Center Buildout

The agreement also sees Nvidia invest $1 billion into South Korea’s Naver, a cloud company building data centers around its GPUs. This move aims to secure AI computing capacity ahead of the 200-megawatt capacity rollout.

SK Group’s $750 Billion AI Infrastructure Push

SK Group, South Korea’s largest conglomerate, pledged $750 billion in AI infrastructure investments, including a partnership with Nvidia valued at over $500 billion. The initiative spans next-generation memory development and large-scale data centers, with SK Telecom’s 2-gigawatt facility as a flagship project. He also revealed discussions with Anthropic, OpenAI, and Broadcom, which are seeking to secure chip supplies for their AI projects.

SK Group’s plans extend beyond memory deals. Samsung Electronics signed a $200 billion memorandum of understanding with Broadcom to expand collaboration on memory chips, foundry technologies, and advanced packaging. This partnership aims to support the next generation of AI infrastructure, with Samsung leveraging its HBM technology and sub-2-nanometer foundry processes. The agreements reflect a broader trend of U.S. AI firms forging closer ties with South Korean semiconductor leaders to navigate supply constraints.

South Korea’s Vision for a Global AI Ecosystem

South Korean President Lee Jae Myung framed the deals as part of a broader dynamic AI ecosystem designed to position the nation as a global AI hub. At the summit, he unveiled the San Francisco AI Declaration, outlining cooperation with U.S. tech companies to expand AI applications in industry and daily life.

A view shows Samsung Electronics
Photo: Reuters

The summit also featured collaborations with academic and research institutions. Nvidia and Korea Advanced Institute of Science and Technology (KAIST) launched a joint AI research lab focused on agentic AI, while Hyundai Motor Group committed to deploying 50,000 NVIDIA Blackwell GPUs for physical AI platforms. These initiatives underscore South Korea’s push to build a full-stack AI ecosystem, from chip manufacturing to infrastructure deployment.

Analysts note the deals could reshape AI supply chains, with South Korean firms gaining critical influence over memory and chip production. However, questions remain about the feasibility of 2027 data center timelines and whether demand will outpace supply.

You may also like