Understanding the Market | Reports say Alibaba and Baidu use self-developed chips to train AI models. Alibaba opened nearly 6% higher, and Baidu opened nearly 4% higher
Alibaba and Baidu opened higher in the morning. As of the time of writing, Alibaba-W rose 5.86% to HKD 151.7; Baidu Group-W rose 3.76% to HKD 110.5. On the news front, according to reports, four informed sources stated that Alibaba and Baidu have begun using internally designed chips to train their AI large models, replacing some NVIDIA chips. It is reported that Alibaba has been applying self-developed chips to the training of lightweight AI models since the beginning of this year, while Baidu is experimenting with the Kunlun chip P800 to train the new version of the Wenxin large model. Notably, overseas OpenAI announced a joint design of its first self-developed AI chip with Broadcom, and Google is accelerating its self-developed TPU to enter third-party data centers to compete directly with NVIDIA. China Merchants Securities pointed out that the rise of self-developed AI chips indicates that the AI infrastructure industry is shifting from "single GPU supply constraints" to "diversified custom chip solutions," with investment logic marginally shifting from hardware monopoly to ecological competition
According to Zhitong Finance APP, Alibaba and Baidu opened higher in the morning. As of the time of publication, Alibaba-W (09988) rose 5.86% to HKD 151.7; Baidu Group-W (09888) rose 3.76% to HKD 110.5.
On the news front, it was reported that four informed sources stated that Alibaba and Baidu have begun using internally designed chips to train their AI large models, replacing some NVIDIA chips. Reports indicate that Alibaba has been applying self-developed chips for training lightweight AI models since the beginning of this year, while Baidu is experimenting with the Kunlun P800 chip to train the new version of the Wenxin large model.
It is worth noting that overseas OpenAI announced a joint design of its first self-developed AI chip with Broadcom, and Google is accelerating the self-development of TPU to enter third-party data centers to compete directly with NVIDIA. China Merchants Securities pointed out that the rise of self-developed AI chips signifies that the AI infrastructure industry is shifting from "single GPU supply constraints" to "diversified custom chip solutions," with investment logic marginally shifting from hardware monopoly to ecological competition
