Sponsored by Google Cloud
Choosing Your First Generative AI Use Cases
To get started with generative AI, first focus on areas that can improve human experiences with information.
Huawei Cloud’s CEO believes Chinese AI developers should focus on improving computing architectures rather than focusing on accessing the most powerful AI hardware
China can maintain a leadership position in AI despite facing restrictions on access to AI hardware, according to Huawei Cloud Computing Technologies CEO Zhang Ping'an.
U.S. companies are banned from exporting high-powered AI chips to China over national security fears. Previous reports suggest some Chinese companies have sought alternative strategies to get around the sanctions and continue AI development.
Speaking at the World Artificial Intelligence Conference (WAIC) in Shanghai, Zhang acknowledged the computational limitations Chinese AI developers face but argued against an overreliance on advanced AI chips.
“If we believe that not having the most advanced AI chips means we will be unable to lead in AI, then we need to abandon this viewpoint,” Zhang said, according to the Times of India.
Huawei offers its own line of AI chips called Ascend. The company claims its latest version, Ascend 910B, matches the computing power and performance of Nvidia's A100 GPU. However, Nvidia has since released more advanced models, including the H100 and Blackwell series.
Huawei's A100 competitor powers the company’s Ascend AI cloud service, which is used by several Chinese companies to train their AI models. Launched last July, the Ascend service provides users with access to computing power, as well as a dedicated platform for training models and algorithms.
Huawei faces difficulties increasing the production of its Ascend chips due to U.S. sanctions limiting access to advanced chip-making equipment, according to Light Reading.
Zhang argued, however, that cloud providers like Huawei could help address the chip shortage for Chinese AI developers. He suggested that cloud solutions could power AI capabilities without relying on the most advanced chips.
The Huawei Cloud CEO advocated for a more integrated approach, emphasizing improvements in computing architectures rather than focusing solely on hardware access.
“We cannot rely solely on having the AI chips with the advanced manufacturing nodes as the ultimate foundation for AI infrastructure,” Zhang said.
Huawei itself operates under U.S. sanctions due to national security concerns.
You May Also Like