CryptoHub network news, Moore Thread-U said at the earnings release meeting that the company's next-generation “Huagang” architecture is under normal development, and the team is pushing the related work with full force. The specific launch timeline will be determined as soon as possible based on the final testing and verification results. Compared to S5000, the company's new generation all-in-one GPU architecture “Huagang” has increased computing power density by 50%, and achieved a tenfold leap in energy efficiency ratio. The Huagang architecture supports full-precision calculations from FP4 to FP64, integrating AI computing acceleration and physical simulation functions on a single chip, enabling the company's GPU products to meet complex requirements from large model training to embodied intelligence, digital twins, and other composite scenarios.