Samsung SDS AI infrastructure developments

Samsung SDS has expanded its AI infrastructure offerings in South Korea through both proprietary cloud services and government-backed research support.

In July 2026, the company launched NPUaaS, a subscription-based cloud platform utilizing Puriosa AI’s second-generation ‘Renegade’ neural-processing unit. This service, delivered via the Samsung Cloud Platform, allows users to scale NPU chips to manage high-performance AI inference without maintaining on-premise hardware. The company positioned the solution as a cost-effective and low-power alternative to traditional GPUs, specifically targeting sovereign-cloud needs for public-sector clients.

By August 2026, Samsung SDS was selected as a primary provider for the ‘2026 AI Research Computing Support Project.’ This government initiative, led by the Ministry of Science and ICT and the Institute of Information & communications Technology Planning & Evaluation (IITP), provides computing resources to domestic academic and research institutions.

With a project scale of approximately 15.3 billion won, Samsung SDS provides high-performance computing resources, including NVIDIA H100 GPUs and the latest B300-based GPU clusters. These B300 units are intended to offer improved performance and larger memory capacity for training and inference. Beyond hardware, the company provides a comprehensive environment including CPUs, memory, storage, and high-speed interconnect networks to reduce the infrastructure burden on researchers. This service is scheduled to run through March 2027.