AMD and NAVER Cloud on Wednesday said they will work together to build AI infrastructure in South Korea. The deal centers on computing hardware and software alignment. NAVER Cloud will expand the use of AMD EPYC processors in its data centers. AMD will give NAVER Cloud early access to its new Instinct MI455X GPUs for testing and evaluation. The partners said they will optimise NAVER Cloud’s AI services on AMD platforms and the ROCm software stack.
Both companies framed the work as technical cooperation. They described engineering tests, software tuning, and joint research to ready AI services for customers. The statement focuses on capabilities and evaluation plans.
“Sovereign AI infrastructure plays a critical role in accelerating how nations build and deploy advanced AI,” said Dr. Lisa Su, chair and CEO, AMD. “Our expanded collaboration with NAVER Cloud brings together AMD’s leadership AI compute platforms and open software ecosystem with NAVER’s cloud and AI capabilities to enable scalable infrastructure and accelerate the next generation of AI in Korea.”
“NAVER Cloud is focused on building scalable and open AI infrastructure to support the rapid growth of AI innovation,” said a NAVER Cloud representative. “Through our collaboration with AMD, we will leverage high-performance AMD compute platforms to enhance our AI and cloud services and deliver new capabilities to customers and developers.”
The timing of the announcement comes as AMD’s chief executive reports activity in South Korea. Lisa Su is on a visit that includes a scheduled tour of Samsung’s chip production site and meetings with Korean partners. Media reporting on that trip links broader supply and partnership discussions, such as memory and foundry supply, to recent industry moves around AI infrastructure.
EPYC processors are server CPUs used for general computing. Instinct MI455X is a data-center GPU model that AMD positions for training and inference tasks. ROCm is AMD’s open software stack for GPU compute. The release says the MI455X units will be provided for evaluation in NAVER Cloud’s cloud and production environments. How many units, and on what timelines, is not in the release.
NAVER Cloud is already a large regional cloud provider. It runs enterprise services and has been active in AI projects in Korea. AMD is a leading supplier of CPUs and GPUs to cloud and edge operators. The announcement is consistent with industry moves to pair silicon vendors with cloud operators to accelerate service validation and customer readiness. But this agreement is a cooperation framework, not a product launch.
Also Read: Inside NVIDIA Corporation’s Strategic Bet on Mira Murati’s AI Startup, Thinking Machines Lab
The companies will optimise software stacks and collaborate on research to enable new AI services. It does not provide technical benchmarks. It does not commit to published performance results or commercial availability dates. Those details matter to customers who plan capacity and budgets for AI workloads. The release leaves those points open.
This collaboration should be read in the context of supply and scale pressures across the AI compute market. Memory such as high-bandwidth memory (HBM) and access to GPUs are strategic inputs for cloud providers building large AI clusters. Industry reporting shows suppliers and cloud operators are seeking closer ties to secure capacity and to tune hardware for specific workloads.
