In the field of artificial intelligence, China has been receiving good news. Since the end of last year, DeepSeek has become popular around the world, and has fought back the assertion that "China cannot make first-class large models" with its strength. Recently, Huawei launched a new model Pangu Ultra MoE with a parameter scale of up to 718 billion. This is a quasi-trillion MoE (mixed expert) model trained on the domestic Ascend AI computing platform throughout the entire process. The practical results prove that domestic computing power can also train the world's advanced large models. China and the United States are regarded as the first echelon of artificial intelligence development. Compared with China, the United States started earlier in the field of artificial intelligence and has more advantages in core technology, capital investment and ecological maturity. Since the debut of ChatGPT and Sora, some voices believe that the gap between China and the United States in artificial intelligence is getting bigger and bigger, and it is difficult to catch up. In fact, the competition between China and the United States in the field of artificial intelligence presents a dynamic change of "the United States is strong and China is fast". China is catching up quickly with application innovation, data scale and policy drive, and the gap continues to narrow. The success of DeepSeek has proved that China has embarked on an innovative path of "low cost and high performance" in the field of large models.