Sharp Partners with Foxconn to Enter AI Server Market by 2027

Sharp announced its entry into the AI server market through a partnership with Foxconn, targeting sales starting in 2027 for enterprise, government, and research customers. The company plans to release modular enterprise servers equipped with NVIDIA RTX PRO 6000 Blackwell GPUs as well as liquid-cooled servers powered by the NVIDIA HGX Rubin NVL8 platform. This expansion highlights how hardware manufacturers are leveraging strategic parent relationships to capture market share in high-demand AI infrastructure. By designing servers around future architectures like NVIDIA Rubin and Blackwell, Sharp aims to serve both localized enterprise deployment needs and massive large language model (LLM) training workloads. The enterprise server features dual Intel Xeon 6 processors, up to 3TB DDR5 RDIMM memory, 38.4TB storage, four 3200W power supplies, and up to 8 NVIDIA RTX PRO 6000 Blackwell GPUs. The LLM-focused server utilizes liquid cooling, interconnects 8 GPUs via high-speed NVLink, and supports NVIDIA ConnectX-9 SuperNICs alongside BlueField-4 DPUs for advanced networking and data processing.