Relevant sources from Zhongke Shuguang revealed that the company will release a next-generation term acceleration plan for large model inference, focusing on dealing with repetitive computation and data scheduling issues in the inference process. At present, Zhongke Shuguang Parastor distributed storage has been applied to Shuguang 8000, the first nationally produced 100,000-card AI hypercluster in China, to provide data throughput and flexible expansion support for large-scale computing tasks. According to the latest report from CCD Consultants, Zhongke Shuguang ranked first in the four industry segments of AI, education, physical intelligence, and autonomous driving.

Relevant sources from Zhongke Shuguang revealed that the company will release a next-generation term acceleration plan for large model inference, focusing on dealing with repetitive computation and data scheduling issues in the inference process. At present, Zhongke Shuguang Parastor distributed storage has been applied to Shuguang 8000, the first nationally produced 100,000-card AI hypercluster in China, to provide data throughput and flexible expansion support for large-scale computing tasks. According to the latest report from CCD Consultants, Zhongke Shuguang ranked first in the four industry segments of AI, education, physical intelligence, and autonomous driving.