Chinese technology giant Alibaba on Wednesday unveiled a new artificial intelligence chip that it said is three times more powerful than its predecessor, as U.S. export restrictions continue to limit Chinese access to advanced processors made by companies such as Nvidia.
Alibaba said its new Zhenwu M890 chip delivers three times the performance of the current Zhenwu 810E processor, underscoring China’s accelerating push to develop domestic AI infrastructure and reduce reliance on foreign semiconductor technology.
According to the company, the new processor features 144 GB GPU memory and interchip bandwidth of 800 GB per second.
The e-commerce and cloud computing giant also disclosed that it had already supplied 560,000 Zhenwu units to more than 400 customers spanning 20 industries, signalling growing adoption of its AI hardware platform within China.
The launch is expected to strengthen Alibaba and its semiconductor subsidiary, T-Head, in China’s increasingly competitive AI processor market, where firms such as Huawei and Cambricon are also expanding their presence.
“Myron Xie, an analyst at SemiAnalysis with a focus on AI accelerators, said: “Alibaba designed AI chips are making headway with external customers and are becoming one of the more popular platforms among Chinese domestic AI hardware chips.”
However, Xie noted that the chip’s advertised memory capacity and bandwidth still trail those of leading Western competitors.
“He added that Alibaba has yet to disclose other critical performance indicators, including compute performance metrics.
Chinese AI developers have for years faced restrictions on purchasing cutting-edge processors from foreign manufacturers, particularly Nvidia, due to US export controls aimed at limiting China’s access to advanced semiconductor technology.
At the same time, Beijing has tightened oversight of domestic companies’ use of foreign AI chips, including Nvidia’s H200 processor, despite Washington recently approving its sale to China.
Alibaba’s latest processor is seen as part of China’s broader effort to build self-sufficient AI computing infrastructure capable of supporting large-scale AI models, including the company’s Qwen series of large language models.
The company also announced on Wednesday that its next-generation AI model, Qwen3.7-Max, would soon be released.
Earlier in April, Alibaba and China Telecom announced plans to launch a data centre in southern China powered by Alibaba’s in-house AI chips, further deepening the company’s investment in domestic AI computing capabilities.
Boluwatife Enome
