News

Enterprise

Artificial Intelligence

Asia

AMD and NAVER Cloud Partner to Build Sovereign AI Infrastructure in South Korea

AMD and NAVER Cloud have entered a strategic alliance to deploy next-generation processors and GPUs for Korea's national AI infrastructure.

AMD and NAVER Cloud have entered a strategic alliance to deploy next-generation processors and GPUs for Korea's national AI infrastructure.

NewDecoded

Published Mar 19, 2026

Mar 19, 2026

3 min read

Image by AMD

AMD and NAVER Cloud announced a strategic collaboration on March 18, 2026, to accelerate the development of sovereign AI infrastructure in South Korea. This partnership aims to integrate AMD's high-performance compute platforms into NAVER’s cloud services to support national-scale AI capabilities. By leveraging local data control and open-source software, the initiative seeks to establish a secure and scalable ecosystem for Korean enterprises as detailed in the official press release.

Under the new agreement, NAVER Cloud will expand its deployment of AMD EPYC processors, including early adoption of the upcoming 6th-Gen Venice lineup. These processors, built on a 2nm process node, are expected to provide up to 256 cores to handle massive AI training workloads. Additionally, NAVER will gain early access to AMD Instinct MI455X GPUs, which feature a significant 432GB of HBM4 memory for enhanced inference performance.

Beyond hardware, the two companies will jointly optimize NAVER’s AI services and software stacks using the AMD ROCm open software platform. This technical alignment is designed to improve the efficiency of HyperCLOVA X, NAVER's proprietary hyperscale large language model. Both organizations also plan to invest in research and development to create new AI solutions tailored specifically for regional customers.

Dr. Lisa Su, chair and CEO of AMD, emphasized the importance of this move for national digital autonomy. "Sovereign AI infrastructure plays a critical role in accelerating how nations build and deploy advanced AI," Su stated during the announcement. NAVER CEO Soo Yeon Choi added that the collaboration leverages high-performance platforms to deliver new capabilities to developers and enhance cloud services.

This alliance reflects a shared goal of providing organizations in Korea with greater control over their data and performance. By reducing reliance on closed ecosystems, the partnership enables a more flexible and competitive landscape for AI innovation. The collaboration further strengthens the existing relationship between the two tech giants in the rapidly evolving Asian market.


Decoded Take

Decoded Take

Decoded Take

This collaboration represents a tactical masterstroke for AMD as it builds a vertically integrated stronghold within South Korea's technology sector. By securing NAVER Cloud as a marquee customer for its 2nm Venice and MI455X hardware, AMD is effectively bypassing traditional supply chain hurdles while challenging NVIDIA's market dominance. The partnership is deeply intertwined with the broader industry shift toward Sovereign AI, where nations prioritize localized data residency and proprietary models like HyperCLOVA X over universal Western solutions. The technical synergy with Samsung Foundry for HBM4 production suggests that AMD has successfully established a localized ecosystem in Korea, spanning from chip manufacturing to cloud deployment.

Share this article

Related Articles