South Korea’s Naver Cloud focused on providing sovereign AI as it eyes global expansion

The South Korean neocloud partnered with Dell Technologies and NVIDIA to drive its mission of realizing Sovereign AI and accelerate expansion into global markets.

Regulators in Asia are continuing to ensure companies embarking on their AI journey meet all data sovereignty requirements. Be it financial services or government agencies or even enterprises, regulators want these organizations to work with the tech companies that can provide such services for them.

This is one of the reasons why neoclouds have seen a surge in Asia. Neoclouds specialize in providing high-performance GPUs (graphics processing units) for AI and HPC workloads. In Asia, neoclouds are a core component to ensure companies have better data management in their AI journey.

NAVER cloud, a South Korean neocloud partnered with Dell Tecnologies and NVIDIA to ensure its HyperCLOVA X omni-model large language model addresses local challenges as the company looks to expand its global impact as well.

Through the Dell AI Factory, NAVER Cloud engineered HyperCLOVA X to drive its mission of realizing Sovereign AI and accelerate expansion into global markets. This includes enabling accelerated AI model training and innovation, strengthened AI and data sovereignty with secure networking for local markets and having enhanced global scalability for multilingual AI solutions.

Dr Kim Yuwon, CEO of Naver Cloud (pictured above with Richard McLaughlin, President, APJC, Dell Technologies) was at Dell Technologies World in Las Vegas to share more about the partnership with Dell and NVIDIA, especially on how it enabled them to truly understand customers' workload and adapting them perfectly.

“Naver is South Korea's number one IT portal. We offer over 200 services like search, math, finance, and more. Naver is one of the few search engines who defended its own home market against the global tech giant, Google. Naver Cloud built and run all infrastructure for these services. And now we are bringing this proven power to the global market,” Dr Kim said.

As a full stack AI infrastructure company, Dr Kim explained that Naver Cloud provides everything to build, train and run AI. This includes operating a huge data center, running multi-hundred megawatts of infrastructure, including tens of megawatts with the latest GPUs.

“Because of these skills, we are leading South Korea's national AI data center project. We also have a strong global network. We have regions in Japan, Singapore, Germany, and the US. And in Korea, we have a large market share in the public sector. And the top companies in Korea really trust us. For example, we power Hyundai Motors with a GPU as a Service, and the Bank of Korea, the central bank, with our private AI cloud. Among these, the GPU as a service is our biggest strength. We run the largest GPU cluster in Korea to supply stable compute power,” he said.

Partnership with Dell Technologies

In the partnership with Dell Technologies and NVIDIA, Dr Kim said that Naver Cloud’s own generative AI model, HyperCLOVA X, is an open source model that allows them to learn and know exactly what infrastructure is needed.

“This experience gives us a huge edge in building an AI factory. With Dell and Nvidia, we can deploy AI environment very fast. Now we are delivering faster, safer, and cheaper AI. Our partnership with Dell is key to our success. Naver Cloud built its AI environment on the Dell AI factory using Dell PowerEdge XE series. When choosing a partnership, raw specs were not the most important thing. We wanted a partner who truly understands AI and can grow with us. And Dell was the perfect match,” he added.

For Dr. Kim, the partnership with Dell brought three clear results. First, it enabled them to train AI faster. Second, it made AI and data sovereignty much safer. And third, it made a strong base for global expansion.

“Infrastructure and models only matter if they solve real world problem, and that is what we do globally. In Japan, our AI calls to elderly people living alone to check their health and safety. In Thailand, we are partnering with local sovereign AI company to develop Thai LLM. In Saudi Arabia, we made a joint venture to build a digital twin and digital map. The AI market is moving fast, but very few companies have data, infrastructure, models, and services all together. Naver Cloud is one of them,” Dr. Kim explained.

Dr. Kim also pointed out that Naver Cloud's biggest advantage is its flexibility. As it has deep experience in building, training, and running AI model, Naver Cloud trully understands customers' workload and can adapt to them perfectly. He also highlighted that in a world where sovereignty is key, Naver Cloud is highly flexible within each customer's specific security and governance needs.

“Our goal is to be a very flexible and powerful AI cloud solution to the world,” he said.

“We want to provide customers that prioritize sovereignty, regulations, and security. And we want to offer them with dedicated cloud and private cloud. And on top of that, we want to integrate with AI technology. So, we want to provide these solutions globally. And we want to reach all the customers in the global market who want these kinds of solutions. And ultimately, what we want to create is the unprecedented vertical that does not exist in the world today,” Dr. Kim concluded.