HANGZHOU, China | Alibaba Group has unveiled a new artificial intelligence chip and announced plans to develop a model with as many as 10 trillion parameters, expanding its push to build a full domestic AI technology ecosystem as Chinese companies face tighter U.S. restrictions on advanced processors.
The company introduced the Zhenwu V900 at Alibaba Cloud’s annual Apsara Conference in Hangzhou on Tuesday. Alibaba said the chip is three times more powerful than its predecessor and described it as the most powerful AI chip in China. The company said the processors can be connected in clusters of up to 500,000 chips for training and operating large AI models.
Alibaba Chief Executive Eddie Wu said the company’s Qwen team plans to develop models containing between 5 trillion and 10 trillion parameters. Alibaba’s current flagship Qwen 3.8 Max has about 2.4 trillion parameters, according to Reuters. The company said its next-generation Qwen 4 is already being trained, with future models expected to scale further.
Alibaba is also planning a major expansion of its computing infrastructure. The company said Alibaba Cloud’s global data center capacity is expected to exceed 20 gigawatts by 2032, reflecting the enormous computing requirements of increasingly sophisticated AI systems. Reuters reported that Alibaba has committed more than $53 billion over three years toward AI-related infrastructure and development.
The announcements come as Chinese technology companies accelerate efforts to develop alternatives to Nvidia’s advanced AI processors. U.S. export controls have restricted Chinese access to some high-end semiconductor technology, encouraging companies including Alibaba and Huawei to invest more heavily in domestic chip development. Reuters reported that Alibaba’s earlier Zhenwu M890 chip, introduced in May, was designed for emerging AI-agent workloads and formed part of a broader semiconductor roadmap.
Alibaba’s latest strategy links chips, AI models and cloud infrastructure rather than treating them as separate businesses. The company has increasingly positioned its Qwen model family as a major part of its cloud and AI services, while developing its own semiconductor technology through its T-Head unit. Alibaba’s financial filings said AI-related products accounted for 30% of external Cloud Intelligence Group revenue in the final quarter of fiscal 2026.
The timing also places Alibaba’s announcement within the wider U.S.-China technology competition. The company’s new AI roadmap was unveiled shortly before a planned meeting between U.S. President Donald Trump and Chinese President Xi Jinping, where technology and AI are expected to be among the issues discussed.
Alibaba’s push follows a broader transformation in China’s AI sector. The country has sought to strengthen its domestic semiconductor and AI capabilities while facing restrictions on access to the most advanced U.S.-designed processors. Chinese companies have responded by developing their own chips, models and computing infrastructure.
Alibaba was founded in 1999 by Jack Ma and initially became one of China’s largest e-commerce companies. It later expanded into cloud computing, digital payments, logistics and other technology businesses. Alibaba Cloud was established in 2009 and has since become a central part of the group’s strategy as the company shifts more heavily toward AI and cloud services.
The latest announcements mark another step in that transformation, with Alibaba seeking greater control over the hardware, models and computing infrastructure needed to compete in the rapidly expanding global AI industry.
Carlo Juancho FuntanillaFrontend Developer, WordPress, Shopify
Contributing Editor
AMA ACLC San Pablo
- Carlo Juancho Funtanilla
- Carlo Juancho Funtanilla
- Carlo Juancho Funtanilla
- Carlo Juancho Funtanilla
