Samsung SDS is fostering its AI infrastructure business as a new growth engine, leveraging the solid growth momentum of its cloud business. Moving beyond enterprise AI transformation, the company is expanding its business footprint to encompass graphics processing unit (GPU) and data center construction and operation, strengthening its 'AI full-stack' framework.

Q2 Performance: Cloud Drives Growth, External Revenue Surges 75%

According to the preliminary Q2 2026 earnings announced by Samsung SDS on the 30th of last month, the company recorded revenue of KRW 3.7178 trillion and operating profit of KRW 231.8 billion. Revenue increased by 5.9% and operating profit by 0.7% year-on-year, respectively.

IT service segment revenue reached KRW 1.7625 trillion, up 5% year-on-year. Within the segment, cloud business revenue grew 17% to KRW 779.4 billion, with external cloud revenue from non-Samsung Group affiliates surging 75% and driving overall growth.

The CSP business, which provides services based on the Samsung Cloud Platform, increased by 24%, driven by expanding GPUaaS demand from public institutions and external enterprises. The MSP business, which supports clients' cloud transition and operations, also grew by 17%, fueled by AI transformation in the financial sector and ERP expansion in the shipbuilding industry.

National AI Computing Center Breaks Ground... Samsung SDS Leads Consortium

The National AI Computing Center has emerged as a core pillar of the AI infrastructure business. According to the Electronic Times and Yonhap News, the Ministry of Science and ICT held a groundbreaking ceremony on the 3rd at the Solasido Data Center Park site in Haenam, South Jeolla Province. A total of KRW 2.5 trillion will be invested in the center, with a target completion and operation date of 2028 and a plan to secure more than 15,000 advanced AI semiconductors.

Samsung SDS participated in the project by leading a consortium that includes Naver Cloud, Samsung C&T, Samsung Electronics, Kakao, and KT. The construction and operation of the center will be handled by a public-private joint venture, Korea AI Computing Center and KOAK. According to the Electronic Times report, Samsung SDS is the largest shareholder of KOAK with a 30% stake, while the government holds 29% and Naver Cloud holds 26.1%.

KOAK is currently building power and cooling facilities necessary for operating high-performance AI servers alongside the center's building. Ahead of the center's completion, some AI computing resources will be supplied early in 2027 through Samsung SDS data centers to proactively address the AI computing needs of domestic companies, universities, and research institutions.

Launch of GPU and NPU Services Expands AI Semiconductor Options

Samsung SDS is also expanding its AI semiconductor offerings provided within its own cloud infrastructure. According to an official company announcement, Samsung SDS launched an Nvidia B300-based GPU service through the Samsung Cloud Platform and is also offering NPUaaS utilizing FuriosaAI's second-generation neural processing unit (NPU), 'Renegade.' Enterprises and research institutions can utilize the necessary computing resources on a subscription basis without having to purchase expensive GPU and NPU servers directly or build their own data centers.

Dongtan Data Center Construction in Full Swing... AI Full-Stack Strategy Accelerates

The company's own data center construction is also taking shape. According to a D.elec report, Samsung SDS signed supply agreements worth a total of KRW 750 billion last month with CE Lab and Data Solution regarding the Dongtan data center project, ramping up its AI computing infrastructure procurement.

Samsung SDS is supporting its clients' AI transformation, spearheaded by its enterprise generative AI platform 'Fabrix' and workplace collaboration solution 'Britty Copilot.' By combining GPUaaS and NPUaaS with its data center construction and operation capabilities, the company aims to establish a business system where enterprises can access everything from AI adoption to operation in one place.

In the logistics sector, the company continues its digital platform-centric transformation. It is expanding services that apply AI technology to 'Cello Square,' which supports global logistics execution and supply chain management, to analyze shipping routes, costs, and supply chain risks.

Future Challenges: Turning Infrastructure Investment into Profit

Samsung SDS's future challenge lies in converting large-scale AI infrastructure investments into stable revenue and profit. The construction timelines for the National AI Computing Center and the Dongtan data center, the expansion of GPU and NPU service clients, and the profitability of the external cloud business are expected to serve as key benchmarks that will determine future earnings trends.

[This article is for reference purposes for investment decisions and does not recommend the purchase or sale of any specific stock. The responsibility for stock investments lies with the investor. This article was written with AI assistance.]