NVIDIA Forges South Korean Mega-Alliance: Partnering with SK Hynix, NAVER, LG, and Doosan to Build Gigawatt-Scale AI InfrastructureNVIDIA has officially announced a series of sweeping strategic partnerships with South Korea’s leading technology and industrial conglomerates including NAVER, SK Hynix, LG, and the Doosan Group to construct next-generation AI data centers and advance localized AI ecosystems. While the definitive financial terms for individual agreements remain undisclosed, the combined initiatives represent a massive infrastructure push in the Asia-Pacific region.
The core hardware alliance is anchored by a multi-year agreement with semiconductor pioneer SK Hynix. The memory giant will co-develop next-generation high-bandwidth memory (HBM) architectures tailored specifically to power NVIDIA’s long-term hardware roadmap, spanning high-density data center servers down to consumer-facing RTX Spark PC platforms. In tandem, SK Hynix will integrate the NVIDIA Omniverse platform into its fabrication facilities to create digital twins of its manufacturing lines, optimizing wafer yields and chip-production workflows.
Expanding its footprint with the SK Group, NVIDIA has partnered with telecom titan SK Telecom to build a gigawatt-scale AI cloud data center in South Korea. Built upon the cloud-native NVIDIA DSX platform, the massive computing facility is projected to go live by 2027.
Furthermore, NVIDIA is collaborating with internet giant NAVER to aggressively scale up existing AI data centers, transitioning them from megawatt capacities to gigawatt-scale infrastructure. This expansion is engineered to handle hyperscale workloads for both domestic Korean enterprises and international clients. On the utility front, energy conglomerate Doosan Group has stepped in to co-develop smart, high-efficiency energy management and cooling systems optimized specifically to sustain the heavy thermal demands of NVIDIA’s AI clusters.
Finally, NVIDIA has established a multi-faceted partnership with the LG Group. The broad collaboration spans the development of autonomous robotics and advanced Physical AI frameworks, the deployment of dedicated AI data centers utilizing the NVIDIA DSX platform, and the integration of next-generation autonomous driving software architectures for automotive applications.
Scaling from megawatts (MW) to gigawatts (GW) of infrastructure is like building "dedicated power plants for AI." Because next-generation processors consume a lot of power, a single data center running at gigawatt-level capacity means it will have enough processing power to run multiple large-scale, national-level AI models simultaneously. This is why NVIDIA needed to partner with Doosan Group (a South Korean expert in energy and nuclear engineering) to design the dedicated power supply and cooling systems.
The term "Physical AI," which NVIDIA is developing with LG, represents a strong technology trend. It's a transition from digital AI (AI confined to computer screens) to something with a "physical body" that understands the real world. For example, it could be integrated into industrial robots in factories or autonomous driving systems in cars, allowing AI to accurately and safely see, estimate distances, and move objects in the physical world through NVIDIA's processing power.
Another business perspective is that this deal reflects South Korea's Sovereign AI policy, an effort to build its own AI infrastructure without relying solely on US cloud services (such as AWS or Google Cloud). By partnering with hardware giant NVIDIA, and leveraging NAVER's local software and SK Telecom's network, South Korea will become one of the most influential AI hubs in Asia.
Samsung Health Gets a Massive Overhaul Shifting to Recovery Metrics Ahead of Next-Gen Galaxy Watch.
Source: Reuters
NVIDIA Forges South Korean Mega-Alliance: Partnering with SK Hynix, NAVER, LG, and Doosan to Build Gigawatt-Scale AI InfrastructureNVIDIA has officially announced a series of sweeping strategic partnerships with South Korea’s leading technology and industrial conglomerates including NAVER, SK Hynix, LG, and the Doosan Group to construct next-generation AI data centers and advance localized AI ecosystems. While the definitive financial terms for individual agreements remain undisclosed, the combined initiatives represent a massive infrastructure push in the Asia-Pacific region.
The core hardware alliance is anchored by a multi-year agreement with semiconductor pioneer SK Hynix. The memory giant will co-develop next-generation high-bandwidth memory (HBM) architectures tailored specifically to power NVIDIA’s long-term hardware roadmap, spanning high-density data center servers down to consumer-facing RTX Spark PC platforms. In tandem, SK Hynix will integrate the NVIDIA Omniverse platform into its fabrication facilities to create digital twins of its manufacturing lines, optimizing wafer yields and chip-production workflows.
Expanding its footprint with the SK Group, NVIDIA has partnered with telecom titan SK Telecom to build a gigawatt-scale AI cloud data center in South Korea. Built upon the cloud-native NVIDIA DSX platform, the massive computing facility is projected to go live by 2027.
Furthermore, NVIDIA is collaborating with internet giant NAVER to aggressively scale up existing AI data centers, transitioning them from megawatt capacities to gigawatt-scale infrastructure. This expansion is engineered to handle hyperscale workloads for both domestic Korean enterprises and international clients. On the utility front, energy conglomerate Doosan Group has stepped in to co-develop smart, high-efficiency energy management and cooling systems optimized specifically to sustain the heavy thermal demands of NVIDIA’s AI clusters.
Finally, NVIDIA has established a multi-faceted partnership with the LG Group. The broad collaboration spans the development of autonomous robotics and advanced Physical AI frameworks, the deployment of dedicated AI data centers utilizing the NVIDIA DSX platform, and the integration of next-generation autonomous driving software architectures for automotive applications.
Scaling from megawatts (MW) to gigawatts (GW) of infrastructure is like building "dedicated power plants for AI." Because next-generation processors consume a lot of power, a single data center running at gigawatt-level capacity means it will have enough processing power to run multiple large-scale, national-level AI models simultaneously. This is why NVIDIA needed to partner with Doosan Group (a South Korean expert in energy and nuclear engineering) to design the dedicated power supply and cooling systems.
The term "Physical AI," which NVIDIA is developing with LG, represents a strong technology trend. It's a transition from digital AI (AI confined to computer screens) to something with a "physical body" that understands the real world. For example, it could be integrated into industrial robots in factories or autonomous driving systems in cars, allowing AI to accurately and safely see, estimate distances, and move objects in the physical world through NVIDIA's processing power.
Another business perspective is that this deal reflects South Korea's Sovereign AI policy, an effort to build its own AI infrastructure without relying solely on US cloud services (such as AWS or Google Cloud). By partnering with hardware giant NVIDIA, and leveraging NAVER's local software and SK Telecom's network, South Korea will become one of the most influential AI hubs in Asia.
Samsung Health Gets a Massive Overhaul Shifting to Recovery Metrics Ahead of Next-Gen Galaxy Watch.
Source: Reuters
Comments
Post a Comment