Alibaba unveils Zhenwu M890 chip and Qwen 3.7-Max AI model
By Axel Miller | 20 May 2026
Summary
- AI hardware expansion: Alibaba Group has unveiled the Zhenwu M890 AI accelerator chip through its semiconductor arm T-Head, claiming three times the performance of the previous 810E generation.
- Agentic AI focus: The new chip is designed for “agentic AI” workloads, featuring 144GB memory capacity and high-speed interconnect architecture aimed at large-scale inference and autonomous reasoning systems.
- Qwen ecosystem push: Alibaba also introduced Qwen 3.7-Max, a flagship AI model optimized for long-duration reasoning and coding tasks as the company deepens its domestic AI infrastructure strategy.
BEIJING, May 20, 2026 — Alibaba Group has expanded its artificial intelligence ambitions with the launch of a new in-house AI accelerator chip and an upgraded large language model, reinforcing China’s push to reduce dependence on foreign semiconductor technologies amid tightening U.S. export restrictions.
The announcements were made during the annual Alibaba Cloud Summit in Hangzhou, where the company introduced the Zhenwu M890 AI chip and the Qwen 3.7-Max model as part of a broader strategy focused on “agentic AI” systems capable of handling complex, multi-step tasks with limited human supervision.
The Zhenwu M890 rollout
Developed by Alibaba’s semiconductor subsidiary T-Head, the Zhenwu M890 is positioned as the successor to the Zhenwu 810E chip series.
According to the company, the processor delivers nearly three times the performance of its predecessor and is optimized for AI inference and training applications requiring large memory capacity and fast communication between accelerators. The chip includes 144GB of memory and supports high-bandwidth inter-chip connectivity aimed at enterprise AI clusters.
Alibaba also unveiled the Panjiu AL128 server architecture, which integrates 128 Zhenwu accelerators into a single rack-scale deployment for enterprise customers. The system is being offered through Alibaba Cloud’s Bailian platform in China.
The company further outlined a long-term roadmap, stating that successor chips named V900 and J900 are planned for release in 2027 and 2028 respectively.
Qwen 3.7-Max and the AI platform strategy
Alongside the hardware launch, Alibaba introduced Qwen 3.7-Max, the newest version of its flagship Qwen large language model family.
The company said the model is engineered for advanced coding, reasoning and long-running AI agent tasks, with support for extended operation without major performance degradation.
Alibaba is also expanding its AI ecosystem by integrating third-party Chinese AI models into its Bailian platform, positioning Alibaba Cloud as a broader infrastructure layer for enterprise AI deployment across industries.
The launch aligns with Alibaba’s previously announced commitment to invest more than 380 billion yuan ($53 billion) into cloud and AI infrastructure over three years.
Why this matters
- China’s AI self-sufficiency drive: The launch highlights China’s accelerating efforts to build domestic alternatives to advanced AI chips from companies such as Nvidia amid escalating export restrictions from the United States.
- Vertical AI integration: By developing both AI chips and foundation models internally, Alibaba is building a vertically integrated AI stack that combines cloud infrastructure, semiconductors and enterprise software services.
- Enterprise AI competition: The Panjiu AL128 architecture and Qwen ecosystem expansion place Alibaba in more direct competition with Chinese rivals including Huawei and domestic AI startups pursuing large-scale enterprise AI deployments.
FAQs
Q1. What is the Zhenwu M890 chip?
The Zhenwu M890 is Alibaba’s newest AI accelerator chip developed by T-Head for AI training and inference workloads, particularly applications involving autonomous AI agents and large-scale enterprise AI systems.
Q2. What is Qwen 3.7-Max?
Qwen 3.7-Max is Alibaba’s latest flagship large language model focused on advanced reasoning, coding and long-duration AI agent tasks.
Q3. What is the Panjiu AL128 system?
Panjiu AL128 is Alibaba’s server architecture that combines 128 Zhenwu M890 accelerators into a rack-scale AI computing platform for enterprise deployments.
Q4. Why is Alibaba investing heavily in AI infrastructure?
Alibaba is expanding its AI infrastructure to support growing enterprise demand while reducing reliance on foreign chip suppliers amid ongoing U.S.-China technology restrictions.