In sum – what you need to know
Enterprise AI – Fujitsu preps new service to run generative AI without companies needing to own or manage their own gear; incorporates Supermicro OEM GPU-accelerated hardware, and Fujitsu’s LLM for enterprises.
All-in platform – the managed platform bundles top-tier AI hardware, support, and management tools into a one-stop enterprise AI shop, removing barriers to enterprise AI adoption (the theory goes).
Notable because – it delivers high-end generative AI for enterprises in a dedicated, compliant environment – key for industries handling sensitive data – with a unique focus on Japanese-language AI for cross-border business.
Fujitsu Limited is to launch a new managed service platform for enterprise customers to run generative AI workloads securely and at scale, without the need to invest in or manage their own infrastructure. The proposition has been developed with US-based server specialist Super Micro Computer (Supermicro), and built around its high-performance, GPU-accelerated server hardware. It also wraps in Fujitsu’s Japanese-language large language model (LLM), Takane, pitched to international enterprise customers looking to do business in Japan.
Takane is part of Fujitsu’s Kozuchi AI suite, launched a year ago as a horizontal platform of AI technology services, but has since morphed into a broader AI development and consultancy proposition; it is being offered on a subscription, via its “all-in-one” ‘data intelligence’ platform as-a-service (DI PaaS), which forms part of its Uvance consultancy portfolio of vertical solutions for “cross-industry societal challenges”. The new AI service platform includes GX2570 M8s model PRIMERGY servers from Fujitsu, which are built on Supermicro’s OEM hardware.
Available in two cooling configurations – a 10U air-cooled model and a 4U liquid-cooled version – the servers feature NVIDIA’s HGX B200 GPU for large-scale generative AI workloads. The service is tailored for security and compliance concerns around sensitive data, offering a dedicated environment for AI training and inference. It also includes Fujitsu’s SupportDesk, billed as the world’s first maintenance service for Supermicro’s liquid-cooled servers, backed by a network of 4,000 engineers in Japan, with targeted on-site response times of two hours.
It also comes with a management tool (‘Infrastructure Manager’) both Supermicro and Fujitsu’s PRIMERGY line of x86 servers, allowing centralized deployment, monitoring, and updates. The proposition will be delivered as a full-service enterprise package, combining generative AI hardware and software from July (2025).
Kazushi Koga, corporate executive officer at Fujitsu, said “Fujitsu has created a comprehensive managed service offering that combines OEM servers with high-performance GPUs for generative AI in dedicated environments, along with high-quality maintenance, integrated management services, and the Takane LLM for enterprises… This initiative is an innovative step toward promoting safe and secure generative AI utilization and will contribute to the expansion of our customers’ businesses.”
Charles Liang, president and chief executive at Supermicro, said: “Supermicro is excited to continue its strong relationship with Fujitsu. We believe the combination of Supermicro’s 10U air-cooled model and a 4U liquid-cooled model, both featuring the advanced NVIDIA HGX B200 GPU family, with Fujitsu’s recognized global service offer customers a proven AI solution for the acceleration of AI in the enterprise.”
