Samsung SDS is targeting a March 2027 start for cloud service using Nvidia’s Vera Rubin GPUs, building AI computing supply ahead of South Korea’s planned national AI computing center, The Korea Herald reported.
CEO Lee Joon-hee outlined the target on the company’s Oct. 29 earnings call, the newspaper said. Lee said Samsung SDS aims to be Asia’s first provider of the service, as demand for computing power rises. Google Cloud and Nvidia announced Vera Rubin infrastructure availability on Oct. 21, a development covered in EconoTimes’ earlier story.
Samsung SDS posted third-quarter revenue of 3.77 trillion won, 11.2% higher than a year earlier. Operating profit rose 3.7% to 241 billion won, The Korea Herald reported. The company is targeting operating profit above 1 trillion won ($748 million) in 2027, the newspaper said.
By 2031, Samsung SDS wants its AI infrastructure footprint to exceed 800 megawatts. Its planned approaches include direct investment, equity stakes and designing, building and operating data centers for customers.
Lee said the Ministry of Science and ICT selected Samsung SDS to develop and operate GPU infrastructure. The company plans to begin offering B300 GPU-as-a-service at the west wing of its Dongtan data center in Hwaseong before the end of 2026. It is also preparing a temporary service there for April 2027 to meet computing demand ahead of the national AI computing center’s scheduled 2028 opening in Haenam, South Jeolla Province.
