Nvidia CEO Jensen Huang warns that AI models like DeepSeek V4 running on Huawei chips could impact U.S. leadership in artificial intelligence, amid U.S.-China tech rivalry.
© RusPhotoBank
Nvidia CEO Jensen Huang has issued a warning about potential shifts in the global artificial intelligence market. He noted that if future AI models run entirely on Huawei chips, it could significantly impact the United States' position in this field.
Particular attention is drawn to the anticipated DeepSeek V4 model, which may launch soon. Rumors suggest it can operate on Huawei's Ascend series chips, marking a notable departure from the traditional use of Nvidia GPUs. However, the model might still retain partial support for the Blackwell architecture.
If the shift to Chinese solutions becomes complete, this could be the first instance where a major AI model is deployed outside the American technological ecosystem.
Although Huawei chips currently lag behind their American counterparts in power, Chinese companies are betting on a different approach: scaling and optimization. Using a large number of processors simultaneously allows for high performance without needing the most advanced components.
China maintains additional advantages in resources, including a significant number of AI specialists and access to cheaper energy. This enables the deployment of large-scale computing systems to compensate for hardware limitations.
This situation unfolds against the backdrop of intensifying technological rivalry between the U.S. and China. American authorities continue to discuss tightening restrictions on Chinese companies, though there are occasional signs of policy easing.
Overall, competition in AI is increasingly shifting from hardware to software solutions, efficiency, and ecosystem control. The success of projects like DeepSeek V4 based on Huawei technologies could substantially alter the market dynamics in the coming years.