Alibaba is planning a new generation of AI models with between 5 trillion and 10 trillion parameters, while launching the Zhenwu V900 processor to support increasingly demanding AI workloads.
The announcements were made at Alibaba Cloud’s annual Apsara Conference in Hangzhou. CEO Eddie Wu said the company’s Qwen team is training Qwen 4, with future models including Qwen 4.5 and Qwen 5 expected to scale well beyond the current Qwen 3.8 Max, which has around 2.4 trillion parameters.
Alibaba said the larger models are being designed for complex tasks requiring longer reasoning and execution. The company also reported progress on systems that can identify weaknesses, run experiments and generate training data with limited human involvement.
The Zhenwu V900 was developed by Alibaba’s T-Head semiconductor unit. According to the company, it delivers roughly three times the performance of its predecessor, the M890, and can be connected in clusters of up to 500,000 chips.
The chip launch reflects Chinese technology companies’ broader push to develop domestic alternatives to advanced foreign AI processors. Alibaba is expanding across the AI stack, including foundation models, semiconductors, data centers and computing infrastructure.
Source: EntArabi
