South Korea is pairing national AI policy with large-scale investments in data centers, advanced semiconductors, and GPU computing capacity. According to South Korea's Ministry of Science and ICT (MSIT), the government aims to secure 52,000 high-performance GPUs by 2028 and expand that figure to 260,000 GPUs by 2030 as part of its national AI competitiveness strategy.

Here are the top 5 upcoming data center projects in South Korea driving this expansion: the Fir Hills AI Supercluster (Jeollanam-do AI Data Center), SGC Energy / OCI Gunsan AI Data Center, Naver GAK Sejong Campus (Cloud Ring Phases 1-6), Saemangeum AI Data Center, and the Samsung SDS Tri-Campus AI Data Centers. 

This article highlights the top five largest data center projects in South Korea in 2026, examining their planned capacity, investment commitments, development status, and strategic importance.

List of Top 5 Largest Data Center Projects in South Korea - 2026

Fir Hills AI Supercluster, Jeollanam-do

The Fir Hills AI Supercluster is set to become the largest upcoming data center in South Korea and one of the largest AI infrastructure developments announced worldwide. The project is being developed in Haenam County, Jeollanam-do Province, by Fir Hills Inc., a subsidiary of Stock Farm Road (SFR), in partnership with the Jeollanam-do provincial government, Haenam County, and local development entities. 

In February 2025, the parties signed a USD 35 billion agreement in California, formalizing plans for the AI-focused campus and advancing an earlier memorandum of understanding into a binding development framework.

Capacity and Technical Specifications

The planned facility will deliver more than 3 GW of data center capacity, making it larger than many of the world's existing AI clusters. Developers intend to build large-scale AI computing infrastructure designed for training and running advanced AI models. The campus will feature advanced data center cooling systems, extensive energy storage capacity, and power infrastructure capable of handling the sharp fluctuations associated with AI workloads. 

Investment and Financing

Fir Hills estimates the project will require between USD 10 billion and USD 35 billion in investment, while local government announcements place long-term investment commitments at approximately 120 trillion won through multiple phases extending to 2030. 

Development Status and Timeline

The project remains in the early development stage following the signing of the February 2025 agreement. Construction is scheduled to begin in winter 2025, and developers are targeting 2028 for the completion of the initial major build-out. 

Provincial authorities have committed to accelerating supporting infrastructure, including power grid upgrades, a new 154 kV substation, telecommunications networks, and water supply facilities required for the campus.

Strategic Importance

The AI supercluster represents a major effort by South Korea to expand AI infrastructure beyond the Seoul metropolitan region. The project is expected to create around 10,000 jobs and strengthen Jeollanam-do's position as an emerging technology hub. 

Fir Hills has also assembled a network of academic and industry partners, including Stanford University and KENTECH, alongside advisors such as Alphabet Chairman John Hennessy and former European Commission Vice President Neelie Kroes. Their involvement is intended to support research, technology development, and long-term competitiveness in AI and energy systems.

SGC Energy / OCI Gunsan AI Data Center

The SGC Energy / OCI Gunsan AI Data Center marks SGC Energy’s entry into the fast-growing AI infrastructure market and ranks among South Korea’s largest planned AI data center developments. The project is being developed through a partnership between SGC Energy, KT Corporation, and Mirae Asset Securities, combining expertise in energy infrastructure, telecommunications, and project financing. 

Located within Gunsan National Industrial Complex 2 in North Jeolla Province, the facility is designed to support growing demand for AI computing and hyperscale cloud services. The partners formally launched the project in February 2026 through a cooperation agreement covering development, infrastructure construction, and financing.

Capacity and Technical Specifications

The project is planned as a 300 MW AI data center campus that will be built in phases. Development will begin with a 40 MW modular data center, creating a scalable foundation for future expansion. The facility will occupy approximately 115,000 square meters of land within the industrial complex. 

One of the project's key advantages is its coastal location, which enables the use of deep ocean water cooling to improve efficiency and reduce operating costs. The campus will also benefit from access to on-site power generation infrastructure, allowing developers to secure a stable electricity supply while targeting competitive power usage effectiveness (PUE) levels for AI workloads.

Investment, Partners, and Financing

SGC Energy is leading the development alongside KT and Mirae Asset Securities. KT will contribute telecommunications and IT infrastructure expertise, while Mirae Asset Securities will support project financing and capital structuring. SGC Energy and its affiliate SGC Green Power are part of OCI Group, one of South Korea’s major industrial conglomerates. 

Although the companies have not disclosed total project costs, the partnership was established to coordinate investment, infrastructure deployment, and financing across the entire development lifecycle. Company executives have also indicated that several global technology firms have expressed interest in the campus, potentially supporting future expansion and occupancy.

Development Status and Timeline

The project is currently in the planning and pre-construction phase following the partnership announcement in February 2026. Land control has already been secured through SGC Green Power, which owns the development site within Gunsan National Industrial Complex 2. 

Construction of the initial 40 MW phase is scheduled to begin by the end of 2026, with operations expected to start in the first quarter of 2028. Additional phases will be delivered over time until the campus reaches its planned 300 MW capacity.

Naver GAK Sejong Campus (Cloud Ring Phases 1-6)

Naver's GAK Sejong Campus, also known as the Cloud Ring, is one of South Korea's most significant AI and cloud infrastructure developments. Located in Sejong City, the hyperscale data center campus serves as the foundation for Naver Cloud's expanding AI, cloud computing, and large language model operations. 

The facility was designed to support the company's growing demand for AI training and inference workloads, including its HyperCLOVA X platform. Construction of the first phase began in November 2023, and Naver has since accelerated expansion plans to strengthen domestic AI infrastructure and reduce reliance on foreign cloud providers.

Capacity and Technical Specifications

The Cloud Ring is planned as a six-phase hyperscale campus with a maximum power capacity of 270 MW and room for up to 600,000 servers upon full completion. The site spans approximately 290,000 square meters and is expected to provide 65 exabytes of storage capacity. The first three phases will establish a server infrastructure capable of supporting more than 15,000 racks, while later phases will scale the campus to its full build-out. 

Naver is also deploying advanced Direct Liquid Cooling (DLC) systems to accommodate increasingly dense GPU clusters required for AI workloads. The facility has been engineered to withstand earthquakes of up to magnitude 7.0 and includes five floors, three below ground and two above ground.

Investment and Financing

In April 2026, Naver secured a 400 billion won (USD 270 million) loan from South Korea's Public Growth Fund after receiving approval from the Financial Services Commission (FSC). The financing will support the expansion of GAK Sejong and help Naver deploy advanced AI infrastructure, including next-generation GPU clusters. 

Development Status and Timeline

The first phase of GAK Sejong is already operational, while Naver is moving forward with the second and third phases simultaneously. According to current plans, Phase 2 is scheduled for completion in 2028, followed by Phase 3 in 2029. The remaining phases will be developed as demand for AI and cloud services continues to grow. 

Naver has also expanded its AI computing capabilities in parallel, including the deployment of large Nvidia GPU clusters to support advanced AI model training and enterprise cloud services.

Saemangeum AI Data Center

Hyundai Motor Group is developing one of South Korea's most ambitious AI infrastructure projects through its planned AI Innovation Hub in the Saemangeum area of Gunsan City, North Jeolla Province. The initiative forms part of a broader investment program worth approximately KRW 9 trillion (USD 6.3 billion) and is being carried out in partnership with the South Korean government and Jeonbuk State. 

Hyundai signed a memorandum of understanding in February 2026 to establish a large-scale industrial ecosystem that combines AI computing, robotics manufacturing, hydrogen production, renewable energy generation, and smart city technologies. At the center of the development is a dedicated AI data center that Hyundai describes as the "brain" of its future AI ecosystem.

Capacity and Technical Specifications

The AI data center will receive approximately KRW 5.8 trillion (USD 4 billion) in investment and is expected to deploy up to 50,000 GPUs through a phased rollout. Hyundai plans to use the facility to process massive datasets generated across autonomous driving, robotics, logistics, software-defined vehicles, and smart factory operations. 

The campus will also include extensive energy storage infrastructure to support AI model development and industrial applications. While Hyundai has not disclosed the final power capacity of the facility, industry estimates suggest the campus could require several hundred megawatts of power as GPU deployment expands. 

Energy Infrastructure and Financing

A defining feature of the project is Hyundai's effort to pair AI infrastructure with dedicated energy resources. The company plans to build a 200 MW-class PEM electrolyzer plant for large-scale hydrogen production and develop gigawatt-scale solar generation capacity by 2035. These energy assets are intended to support the data center's long-term power requirements while reducing exposure to grid constraints and rising electricity demand. 

Hyundai has stated that it is working with financing partners, including the Korea Development Bank, to structure funding for the project. The broader investment package also includes a robotics manufacturing cluster and the development of an AI Hydrogen Smart City within the Saemangeum region.

Development Status and Timeline

The project remains in the planning stage following the signing of the February 2026 agreement. Hyundai intends to begin construction of the AI data center and associated solar infrastructure in 2027, with completion targeted for 2029. 

Development of the hydrogen production facility will follow a similar timeline, while the robotics manufacturing cluster is expected to begin construction in 2028 and enter operation in 2029. The phased deployment strategy allows Hyundai to scale AI computing capacity in line with demand from its vehicle, robotics, and industrial technology businesses.

Samsung SDS Tri-Campus AI Data Centers

Samsung SDS is building a network of AI-focused data centers across Dongtan, Gumi, and Haenam as part of a broader strategy to expand South Korea's domestic AI infrastructure. The company plans to invest approximately 3.7 trillion won (USD 2.8 billion) across the three campuses, creating a distributed AI computing platform that supports high-performance computing, GPU cloud services, and sovereign AI initiatives. 

The program reflects Samsung SDS's transformation from a traditional IT services provider into a cloud and AI infrastructure company, while also supporting South Korea's push to develop AI capabilities using locally owned and operated infrastructure.

Capacity and Technical Specifications

The three campuses will collectively provide more than 140 MW of AI computing capacity. The Dongtan Campus in Gyeonggi Province serves as Samsung SDS's flagship high-performance computing facility and now offers 40 MW of capacity following the completion of its west building expansion. The Gumi Campus in North Gyeongsang Province is being developed as a 60 MW AI data center and will act as the operational base for Samsung SDS's GPU-as-a-Service business. 

The third site, located in Haenam, will host the National AI Computing Center and is expected to support at least 15,000 advanced GPUs by 2028. While the final power configuration has not been confirmed, industry estimates suggest the facility will require at least 40 MW of capacity during its initial phase.

Investment and Financing

Samsung SDS is leading the overall development program, but financing structures differ across the three campuses. The Dongtan and Gumi facilities are being funded directly by Samsung SDS, which has allocated hundreds of billions of won toward their construction and expansion. The Haenam campus forms part of the government's National AI Computing Center (NAICC) initiative, a project valued at approximately 2.9 trillion won. 

Samsung SDS was selected as the preferred bidder through a consortium that includes Naver Cloud, KT, Kakao, Samsung Electronics, and Samsung C&T. Under the proposed structure, Samsung SDS will hold roughly 30% ownership, while Naver Cloud and the South Korean government will also maintain significant stakes. Funding will be secured through a combination of private capital and policy-backed financing.

Development Status and Timeline

The Dongtan campus has already completed its latest expansion and entered operations, making it the most advanced of the three sites. Construction of the Gumi facility is underway, with investment commitments extending through 2029 as capacity is gradually built out. The Haenam campus remains in the planning and negotiation phase following Samsung SDS's selection as the preferred bidder for the National AI Computing Center project in March 2026. 

Government officials and local authorities are also working to accelerate the support for power infrastructure, including a new 154 kV substation and transmission network. Samsung SDS expects the broader tri-campus strategy to take shape between 2027 and 2029 as all three facilities reach operational maturity.

South Korea Data Center Market - 2026

South Korea's data center industry is undergoing a structural transformation driven by the convergence of government policy, chaebol capital, and hyperscaler demand. The country's existing colocation base stands at approximately 58 to 59 operational facilities, with around 32 concentrated in Seoul alone, a density that has made the capital region the dominant hub for digital infrastructure in Northeast Asia. 

The MSIT-led National AI Computing Center, operating as a public-private special-purpose company, is already distributing its first cohort of GPUs across Naver Cloud, NHN Cloud, and Kakao, with full operations targeted for 2027. 

Meanwhile, NVIDIA's October 2025 press release confirmed that the Korean government, through MSIT, is deploying over 50,000 of its latest Blackwell GPUs across national cloud and IT providers as part of the sovereign AI infrastructure drive, alongside parallel commitments from Samsung (50,000+ GPUs) and SK Group (50,000+ GPUs) under the same announcement.

The geographic spread of new capacity is shifting decisively away from Seoul. Provincial governments in Jeollanam-do, North Jeolla, Gyeonggi, and Ulsan are competing for flagship projects, supported by national policies including direct-PPA reforms, tax incentives for data center investment, and streamlined permitting under the AI Basic Act. 

The law explicitly mandates government support for AI data centers as a component of national infrastructure, giving statutory backing to the investment pipeline that is already under construction.

Other top data center projects in South Korea include:

• SK Group - AWS AI Zone (Mipo National Industrial Complex)

• Digital Edge SEL2 Campus

• DC Connects Seoul01 (Yongin)

• Princeton Digital Group SE1 (Incheon Hyperscale Campus)

• National AI Computing Center (NAICC)

• MAIF4/GABIA - Ansan Data Center

• KAKAO - AI Digital Hub Data Center

• PDG - SE1 Bupyeong Data Center Campus

• PDG - SE1 Data Center Campus

Conclusion

South Korea's largest data center projects reflect a broader national effort to build the infrastructure required for the AI economy. Backed by government policy, major corporate investment, and long-term GPU deployment plans, projects such as the Fir Hills AI Supercluster, Naver GAK Sejong, Hyundai's Saemangeum AI Hub, the SGC Energy Gunsan campus, and Samsung SDS's tri-campus network are set to reshape the country's digital infrastructure landscape. 

As construction progresses over the next several years, these developments will play a central role in strengthening South Korea's position as a leading AI and data center market in the Asia-Pacific region.

Find Latest Data Center Facility Projects in South Korea for Business Opportunities.

Subscribe to our database on Data Center Facility Projects and Tenders in South Korea to get access to reliable and high-quality insights on upcoming, under-construction, and completed Data Center Facility Projects across the world or in your desired geographical location.

Our user-friendly platform provides essential details, timely updates, key stakeholder contact information, and business opportunities tailored for engineering companies, industry professionals, investors, and government agencies.