Zhenwu 810E: Alibaba unveils stronger AI chip and Qwen3.6-Plus

Alibaba disclosed simultaneous upgrades on both hardware and model fronts: the company highlighted a more powerful Zhenwu 810E AI processor developed within its chip unit and released Qwen3.6-Plus, a new iteration of its flagship large language model (LLM).

According to company statements and media reports, the Zhenwu 810E is optimized for large-scale inference and training workloads and has been deployed in multi-card clusters on Alibaba Cloud; initial rollouts include a 10,000‑card cluster serving hundreds of enterprise customers across energy, automotive and research sectors.

The announcements have immediate commercial implications: Alibaba Cloud has adjusted pricing for certain AI compute cards and intelligent storage offerings amid surging demand, and market observers noted near-term share-price and demand volatility. Analysts point to the dual impact on cloud revenue mix and unit economics as clients shift toward agentic AI applications.

These moves fit a broader strategic picture in which Chinese tech firms accelerate domestic chip development to reduce reliance on foreign suppliers amid export restrictions. By integrating in‑house silicon, cloud infrastructure and LLM capabilities, Alibaba aims to capture more of the AI value chain and offer vertically optimized products to enterprise customers.

Market consensus suggests that while Zhenwu‑backed deployments could improve Alibaba Cloud’s competitive positioning over time, execution risks remain—manufacturing scale, software-hardware co‑optimization and sustained enterprise demand will determine the magnitude of benefit. Investors and corporate buyers will closely watch utilization, pricing trends and model performance metrics in coming quarters.