While Nvidia’s chips continue to be used for specific tasks, Ant is increasingly turning to alternatives, including those from AMD and local Chinese manufacturers, for its latest AI model developments
Alibaba's Jack Ma-backed fintech giant Ant Group has made a significant breakthrough in the world of artificial intelligence by drastically reducing AI training costs. By utilizing Chinese-made chips from Alibaba and Huawei, the company has been able to lower AI training costs by 20%, according to industry reports. This development has been hailed as a major achievement, as it not only reduces the expense and time associated with training AI models but also lessens China’s reliance on American tech giants like Nvidia.
Although Ant Group has historically relied on Nvidia chips for AI model training, its recent use of chips from Alibaba and Huawei has shown that Chinese-made alternatives can perform on par with Nvidia’s H800 chips, sources reportedly claim. This shift in strategy comes as Ant Group looks to diversify its hardware options. While Nvidia’s chips continue to be used for specific tasks, Ant is increasingly turning to alternatives, including those from AMD and local Chinese manufacturers, for its latest AI model developments.
Earlier this year, a small Chinese AI startup, DeepSeek, made headlines by achieving a major technical milestone with a large language model that rivals those of OpenAI and Google—without the massive training budgets typically associated with such advancements. DeepSeek’s achievement signals China’s growing capabilities in AI development and its drive to challenge the US in the race for AI supremacy. Faced with US restrictions on access to cutting-edge semiconductors, Chinese companies are finding innovative ways to reduce costs while developing competitive AI models.
China’s AI growth amid US restrictions
Despite the US government’s efforts to slow China’s AI progress by limiting access to advanced semiconductors like those produced by Nvidia, China’s AI sector is continuing to advance. While Nvidia’s high-end chips are restricted, lower-end models are still available for sale in China, allowing companies to continue their AI work. This has spurred a nationwide push for the widespread deployment of AI models in sectors such as healthcare, local government, and other industries.
In a related move, Ant Group has made significant upgrades to its AI solutions for the healthcare industry. The company unveiled a new healthcare-specific AI model, integrated into its "All-in-One Large Model Machines for Healthcare." This model, adopted by hospitals in Beijing, Shanghai, Hangzhou, and Ningbo, is designed to answer medical queries and improve healthcare services. Combining DeepSeek’s R1 and V3 models, Alibaba AI innovation, and Ant’s BaiLing, this AI model aims to revolutionize healthcare services in China.
With these innovations and the growing collaboration, Ant Group is positioning itself as a leader in both AI and the broader technology landscape in China, reducing costs while enhancing the country's technological independence.
