Nvidia says South Korea-based Naver will use Nvidia's DSX tech to build AI factories at “gigawatt scale” to meet global demand for AI services and physical AI
Nvidia (NVDA.O) said on Monday that South Korean internet conglomerate Naver (035420.KS) would use its technology …
Context & Ripple Effects
Naver has already been building toward a larger AI infrastructure role: related coverage identifies it as South Korea’s largest Nvidia-chip buyer and shows it positioning its AI cloud services for countries seeking an alternative to U.S. or Chinese systems.
The announcement also sits within a broader cluster of Nvidia partnerships in South Korea, including planned GPU supply to the government, Samsung, Naver and SK Group, plus Samsung’s own Nvidia-backed AI megafactory effort. Naver’s move extends that buildout from chip procurement to operating large-scale AI capacity.
First-order effects
- Naver becomes a major deployment partner for Nvidia’s AI-factory technology, committing its infrastructure strategy more tightly to Nvidia’s platform for AI services and physical-AI workloads.
- Naver gains a clearer basis to expand AI cloud and related services for global customers, while Nvidia gains another large Korean reference deployment beyond its chip-supply arrangements.
Second-order effects
- Naver’s scale-up raises the competitive bar for regional AI-cloud providers: competing offerings will need to match not only model access but also large-scale inference and deployment capacity.
- The buildout reinforces demand across Korea’s AI infrastructure ecosystem, complementing Nvidia-linked efforts at Samsung and SK Group and increasing the importance of local compute, memory, and industrial-AI integration.
Third-order effects
- If similar projects materialize, South Korea could evolve from a major buyer of AI hardware into a base for operating export-oriented AI infrastructure and physical-AI services.
- The pattern strengthens Nvidia’s role as an end-to-end AI infrastructure partner rather than solely a chip supplier, though the durability of that position will depend on whether customers can turn very large capacity commitments into sustained service demand.
The trend: This is part of the shift from buying AI accelerators to building national and corporate-scale AI factories that package compute, cloud services, and physical-AI deployment together.