Ant Group, a key player in the Alibaba ecosystem, is taking a fresh approach to AI development by integrating semiconductors from both Chinese and American manufacturers. This clever strategy is all about boosting efficiency while cutting down on the time and cost of training AI models. By doing so, Ant Group is steering away from relying solely on a single supplier like Nvidia.
Recently, Ant made headlines with some impressive advancements in AI solutions for healthcare. These innovations are now being used in seven major hospitals and healthcare institutions. It’s a big step forward, especially when you consider how AI can transform patient care and streamline medical processes.
Ant Group is tapping into a broader industry trend known as the ‘mixture of experts.’ This technique allows for the training of AI models with fewer computational demands. A paper released earlier this month highlighted that Ant Group has successfully reduced computing costs by 20% by using more affordable hardware for its models. That’s a significant saving!
Known for operating Alipay, China’s leading mobile payment app, Ant Group has shifted gears from mainly using Nvidia chips to incorporating those from Alibaba and Huawei. This move also includes greater reliance on Advanced Micro Devices and other Chinese alternatives, as reported by Bloomberg. However, when CNBC reached out, Ant chose not to comment further.
The healthcare AI model is a collaborative effort involving DeepSeek’s R1 and V3 models, Alibaba’s Qwen, and Ant’s BaiLing. It’s designed to tackle medical inquiries and enhance patient services, which could be a game-changer for healthcare providers.
Meanwhile, the U.S. continues its efforts to limit China’s access to advanced semiconductors crucial for AI development, allowing only the sale of lower-end Nvidia chips to Chinese entities. This ongoing dynamic adds another layer of complexity to the global tech landscape.
