Sharp and Foxconn launch AI servers with nvidia gpus, set for 2027 release

Sharp, in collaboration with Foxconn, has announced its entry into the AI server market, with new products set to be available for purchase in 2027. The initiative targets enterprise clients, Japanese government agencies, research institutions, and data centers. Two AI server models are being introduced, each tailored to specific use cases. The first server is designed for enterprise AI applications, featuring a modular architecture that supports up to 8 NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. It includes dual Intel Xeon 6 series processors, 3TB DDR5 memory, 38.4TB storage, and four 3200W power supplies. The second server is aimed at large AI model training and inference, utilizing NVIDIA HGX Rubin NVL8 with NVLink technology to integrate 8 GPUs. It also incorporates liquid cooling and supports NVIDIA ConnectX-9 SuperNIC and BlueField-4 DPU configurations.