AMD and NAVER Cloud announced a strategic collaboration on March 18, 2026, aimed at accelerating sovereign AI infrastructure development in South Korea. The deal expands an already existing relationship between the two companies and brings together AMD’s open compute platforms with NAVER’s hyperscale cloud and LLM expertise — creating what both sides are positioning as the backbone of Korea’s next-generation AI capabilities.

AMD × NAVER Cloud: Key Collaboration Details

Aspect	Detail
Processor Expansion	NAVER Cloud expanding AMD EPYC™ deployment, including 6th-Gen “Venice”
GPU Access	Early access to next-gen AMD Instinct™ MI455X GPUs
Software Stack	Joint optimisation on AMD ROCm™ open software ecosystem
R&D Focus	New AI services and solutions for NAVER customers
Strategic Goal	Sovereign AI infrastructure for national-scale deployment in Korea
NAVER LLM Status	World’s third company to develop a hyperscale LLM

Why “Sovereign AI” Is the Key Term Here

Sovereign AI refers to a country’s ability to develop, deploy, and control AI infrastructure using its own data, its own compute, and its own policy framework — rather than depending entirely on US hyperscalers like AWS, Azure, or Google Cloud. For Korea, this matters enormously given the scale of NAVER’s domestic dominance and the government’s push to keep sensitive AI workloads within national boundaries.

NAVER Cloud is an unusually credible player in this space. As a subsidiary of NAVER Corporation — the company behind Korea’s dominant search engine, messaging platform, and one of only three organisations globally to independently develop a hyperscale LLM — it brings both the infrastructure pedigree and the AI research depth that sovereign deployments demand.

AMD’s role is equally deliberate. The 6th-Gen EPYC “Venice” processors expand on NAVER Cloud’s existing EPYC deployment for CPU-heavy AI inference and data processing workloads, while the Instinct MI455X GPUs — provided as early access ahead of general availability — handle the heavy lifting on large-scale AI training. The joint optimisation on AMD’s ROCm open-source software stack is the layer that ties it together, ensuring NAVER’s AI services run efficiently on AMD hardware without being locked into proprietary software dependencies.

This is also part of a broader AMD push to challenge NVIDIA’s dominance in AI infrastructure by building deep, bilateral partnerships with major cloud providers across Asia — a region where sovereign AI investment is accelerating fastest.

For more AI and enterprise tech coverage, check out TechnoSports’ tech section.

FAQs

What is the AMD Instinct MI455X and why does early access matter?

The MI455X is AMD’s next-generation AI GPU for large-scale training workloads; NAVER Cloud gets early access before general availability, giving it a meaningful head start in deploying cutting-edge AI infrastructure.

What makes NAVER Cloud a credible sovereign AI partner?

NAVER is the world’s third company to independently build a hyperscale LLM, giving NAVER Cloud both the AI expertise and infrastructure scale needed for national-level AI deployment in Korea.