Chinese tech giant Baidu’s CEO announced on Friday that the company has successfully “illuminated” a cluster of 30,000 of its self-developed, third generation P800 Kunlun chips, capable of supporting the training of DeepSeek-like models.
Robin Li made the announcement at the company’s annual developer conference, where the Chinese search engine leader provided updates on its artificial intelligence (AI) advancements. “Illuminated” refers to the process of switching on the cluster and preparing it for training tasks.
He stated that the P800 cluster can facilitate the training of models with hundreds of billions of parameters, similar to DeepSeek, or enable a thousand customers to fine-tune models with billions of parameters simultaneously.
Baidu reported that Chinese banks and Internet companies have already adopted the P800 chips.
Li also unveiled Baidu’s latest AI model, Ernie 4.5 Turbo, highlighting its performance matching the industry’s top models in several benchmark tests, showcasing capabilities from coding to linguistic understanding. Additionally, the company launched a new reasoning model named Ernie X1 Turbo and announced its plans to integrate its AI capabilities across its applications, including its cloud drive and content platform Baidu Wenku.
“There are many (AI) models, but it’s apps that rule the world. The application is the king,” Li emphasized. “Without apps, models and chips are worthless.”
This product launch occurs during a period of intensified competition in China’s AI market, where tech companies are shifting their focus from developing foundational models to discovering applications beyond AI chatbots that can attract and retain users.
Baidu was among the first major Chinese companies to make significant investments in AI following the debut of OpenAI’s ChatGPT in 2022. However, its Ernie Bot has faced challenges in gaining significant traction amidst the fierce competition.
