SK has expanded its artificial intelligence (AI) infrastructure partnerships with leading global technology companies in Silicon Valley, reinforcing South Korea's ambition to become a major AI infrastructure hub in the Asia-Pacific region.
On July 24 (local time), SK announced at the K-AI Summit in San Francisco that it had reached strategic agreements with NVIDIA, Microsoft, Anthropic, and Amazon Web Services (AWS) covering AI memory, hyperscale AI factories, and next-generation AI data centers. The partnerships are expected to accelerate SK's recently announced Three Mega Projects while deepening Korea's role in the global AI ecosystem.
"The expansion of these partnerships reflects rapidly growing global demand for AI infrastructure, particularly AI memory," an SK official said. "It also strengthens the execution of Korea's national AI strategy."
NVIDIA alliance expands across AI infrastructure
The centerpiece of the announcements was SK's expanded partnership with NVIDIA.
SK Telecom signed a Letter of Intent (LOI) with NVIDIA to collaborate on AI infrastructure projects valued at more than $500 billion, spanning hyperscale AI factories, next-generation computing systems, and long-term investment cooperation.
Under the agreement, SK Telecom will deploy NVIDIA's DSX AI Factory architecture together with the Vera Rubin AI computing platform powered by SK hynix's next-generation HBM4 memory. Construction of the first AI factories is scheduled to begin in 2027.
The infrastructure will support sovereign AI, agentic AI, physical AI, and enterprise AI services while meeting growing AI demand across Korea and the broader Asia-Pacific region.
SK hynix will also expand its long-term collaboration with NVIDIA on next-generation AI memory. The companies will jointly develop and optimize HBM technologies for increasingly demanding AI workloads, including large language models (LLMs), agentic AI, and physical AI.
"Korea has the world-class networking, data center infrastructure, semiconductor leadership, and industrial base needed to become a global AI powerhouse," said Jensen Huang, founder and CEO of NVIDIA. "Together with SK Telecom and SK hynix, we look forward to building the next generation of AI factories."
SK Group Chairman Chey Tae-won said AI competitiveness will increasingly depend on a nation's ability to generate intelligence rather than simply consume AI services.
"Working with NVIDIA, SK will build world-class AI factories that help transform Korea into a global hub for AI innovation," he said.
Microsoft, Anthropic, and AWS deepen cooperation
SK hynix also announced a long-term memory supply partnership with Microsoft to develop AI server memory optimized for next-generation data centers.
SK Telecom separately signed a memorandum of understanding (MOU) with Anthropic to cooperate on gigawatt-scale AI data centers in Korea, expanding a strategic relationship that began with SK Telecom's investment in the company in 2023.
Meanwhile, SK and AWS agreed to broaden their collaboration beyond the Ulsan AI Data Center project by combining SK Telecom's infrastructure capabilities with AWS's global cloud platform and AI technologies to expand AI computing capacity across Korea.
On the sidelines of the summit, Chairman Chey also met with OpenAI CEO Sam Altman to discuss long-term cooperation on AI infrastructure, including AI memory and large-scale computing facilities.
Chey outlines three priorities for Korea's AI future
Addressing government officials, global technology executives, researchers, and investors at the K-AI Summit, Chairman Chey described SK's Three Mega Projects as practical investments driven by accelerating global demand for AI infrastructure.
He identified three priorities for Korea's AI future:
Building an AI-native nation where every citizen can access AI agents and Korea serves as a global AI testbed;
Reducing AI computing costs through expanded AI memory production and hyperscale AI data centers;
Establishing responsible AI governance that promotes innovation while addressing employment and environmental challenges.
"AI is still in its early stage," Chey said. "Korea should leverage its strengths in AI memory and hyperscale computing infrastructure to become a strategic AI hub connecting North America, Europe, and Asia."
Why it matters
The announcements underscore a broader shift in global AI competition. As the industry moves beyond model development toward large-scale computing infrastructure, competitive advantage is increasingly determined by integrated ecosystems spanning semiconductors, AI memory, cloud platforms, data centers, and energy infrastructure.
By expanding partnerships with NVIDIA, Microsoft, AWS, Anthropic, and OpenAI, SK is positioning itself across the AI infrastructure value chain while reinforcing South Korea's ambition to become a strategic AI infrastructure hub for the Asia-Pacific region.
The next challenge will be execution. Delivering gigawatt-scale AI factories and data centers will require sustained investment, sufficient power capacity, regulatory support, and skilled talent. In the AI era, national competitiveness will depend not only on technological leadership but also on the ability to build and operate the infrastructure that powers the global AI economy.
