Nvidia locks down memory from SK Hynix as part of $500 billion AI deal

Nvidia Announced that it has secured the supply of AI memory from South Korea SK HynixThat’s as chipmakers look to lock in key components of their advanced processors and systems.

Nvidia said the deal, announced late Friday in San Francisco, could be worth $500 billion over several years and includes construction of a large data center expected to go online in 2027.

SK Telecom, a subsidiary of SK Hynix, will build a cloud business using Nvidia’s Vera Rubin system. Nvidia said it is targeting enough capacity to require 2 gigawatts of power, which it said represents a massive boost by hundreds of thousands of graphics processing units.

As the AI ​​boom creates a global memory shortage, NVIDIA is proactively securing the supply of high-bandwidth memory essential for its GPUs and systems. According to analysts, SK Hynix is ​​a leader in HBM production.

“This expansion includes the opportunity to co-develop next-generation SK Hynix AI memory, which helps ensure a stable supply of HBM memory,” Nvidia Enterprise Vice President Raj Mirpuri said in a call with reporters.

The agreement also signals that building large-scale AI infrastructure is moving beyond a few hyperscalers, with foreign governments and large conglomerates starting to get involved.

The deal was announced at an AI summit in San Francisco with South Korean officials, including President Lee Jae-myung.

Nvidia also announced Friday that it will invest $1 billion in Naver, a South Korean cloud company that builds data centers around GPUs. Nvidia said the project will provide potential customers in South Korea and around the world the opportunity to secure AI computing capacity before the 200 megawatt capacity is completed.

SK Hynix, South Korea’s second-highest market capitalization company, was listed on the Nasdaq earlier this month as part of a bid to fund infrastructure development.

Attention: South Korea’s AI boom extends to housing