Starting in July 2025, Fujitsu will combine these OEM products, maintenance services, and integrated management tools with its enterprise large language model (LLM) Takane to offer managed services for building generative AI infrastructures, enabling customers to leverage generative AI without owning the assets.

The use of generative AI in handling confidential or personal information introduces risks such as unintentional learning, data breaches, and compliance with internal data storage policies. As a result, a dedicated environment is often required, demanding significant computing power and diverse specialized expertise. Therefore, there is a demand for servers equipped with high-performance GPUs, as well as support for maintenance and operation, and management tools that facilitate stable server performance.

The PRIMERGY GX2570 M8s is a server designed for large-scale generative AI applications and will be offered in two cooling configurations: an air-cooled 10U model and a liquid-cooled 4U model, both featuring the advanced NVIDIA HGX B200 GPU.

The SupportDesk maintenance and operations support service will offer high-quality support for Supermicro servers provided by Fujitsu Group, including on-site response to addresses in Japan within approximately 2 hours, achieved through a network of 4,000 experienced engineers at 700 engineering locations nationwide.

The integrated Infrastructure Manager tool simplifies server lifecycle management, including deployment, monitoring, and upgrades—a key priority for businesses. It enables integrated operation and management of Supermicro GPU-equipped servers and Fujitsu Group's existing PRIMERGY line of x86 servers, freeing up time for customers to focus on using middleware and applications.

The managed service provided by Fujitsu encompasses these OEM products, as well as integrated maintenance and management services, and generative AI models such as Takane. This allows customers to easily deploy a generative AI infrastructure of the required size without owning the assets. By providing this managed service, Fujitsu will enable customers to quickly deploy a generative AI infrastructure, reduce infrastructure maintenance and management costs, and contribute to the adoption of generative AI and the business transformation of their businesses.