Collaboration expands beyond semiconductors to cover full AI infrastructure
SK Group, Nvidia to jointly develop next-generation AI factory architecture
SK Hynix deepens long-term technology partnership on next-generation memory chips
SK Telecom to build AI data centers and expand cloud business with Nvidia
Chey Tae-won: 'We will now raise cooperation to the group level'
Jensen Huang: 'AI infrastructure is just getting started — the future is very bright'
SK Group and Nvidia are elevating their AI partnership to an entirely new level.
Cooperation between the two companies had previously centered on semiconductors — chiefly SK Hynix's high bandwidth memory (HBM) chips — but will now extend across the full spectrum of AI infrastructure, including the construction and operation of AI factories.
SK Group announced Monday that the two companies also plan to move beyond infrastructure buildout and jointly pursue research to co-develop next-generation AI factory architecture.
The new research and development collaboration will include joint work on computing architecture designed to simultaneously improve GPU and memory performance from the design stage. The two sides plan to establish a joint consultative body to advance that effort.
SK Hynix will strengthen a long-term technology partnership with Nvidia to accelerate semiconductor design and manufacturing, including co-developing next-generation memory chips for global AI factory construction.
The long-term partnership is intended to ensure a stable supply of advanced memory given the lengthy development cycles involved. On that basis, SK Hynix will continue supplying memory aligned with Nvidia's infrastructure roadmap and the growing global demand for AI infrastructure buildout.
Through the partnership, SK Hynix will also enter new AI markets that Nvidia is pioneering — including AI infrastructure, personal AI and physical AI.
The two companies plan to jointly develop memory for the Vera Rubin AI supercomputer, the Vera central processing unit, the RTX Spark PC and the Jetson Thor robotic computing platform.
SK Telecom, meanwhile, will partner with Nvidia to build dedicated AI data centers and expand its cloud business.
SK Telecom said it will develop AI factories — dedicated AI data centers optimized for AI training and inference workloads — based on Nvidia's AI infrastructure platform DSX.
The two companies plan to launch the first such facility in South Korea next year to validate the operating model, then expand the business across Asia. Their long-term goal is to build AI infrastructure at the gigawatt scale.
SK Group Chairman Chey Tae-won, speaking at a joint briefing with Nvidia CEO Jensen Huang at SK Seorin Building in Jongno-gu, Seoul, on Monday morning, said the group would build future AI factories together with Nvidia. "Until now, much of our cooperation has been focused on memory," he said. "From this point on, we will raise that cooperation to the group level."
Chey added that the two companies would develop and share a joint R&D roadmap so they could adapt more quickly to future AI demand.
Huang said he was witnessing enormous global demand for more AI factories and that business was booming. "That is why we formed this partnership," he said. "We are only at the very beginning of building AI infrastructure, and the future is very bright."
gil@heraldcorp.com
jeongwan@heraldcorp.com
ko@heraldcorp.com
