Search
On July 10, during the 2026 Intelligent Computing Application Conference hosted by the IFFO (International Freeform Fabrication Organization), Sugon made a groundbreaking announcement: China's first indigenous 100,000-GPU AI supercluster, named 'Sugon 8000 (Dengfeng)', has been successfully constructed and integrated into the National Supercomputing Internet. This milestone signifies a leap forward in AI infrastructure, transitioning from the 10,000-GPU scale to an unprecedented 100,000-GPU capability. The Sugon 8000 leverages cutting-edge 'superintelligence fusion' technology, accommodating a full spectrum of precision levels from FP64 to INT8. It is designed to cater to a diverse range of applications, including scientific computations, extensive model training, AI-driven inference, and industrial simulations, showcasing end-to-end, fully self-developed prowess.
-
C114 Communication Network -
Communication Home
