Alibaba has introduced a new artificial intelligence processor designed specifically for AI agents, marking a major step in the company’s broader push to build a complete AI ecosystem. Alongside the chip launch, the company also revealed a long-term semiconductor roadmap and an upgraded large language model, signaling that Alibaba is no longer focused only on cloud services or software development. It is now investing heavily across the full AI infrastructure stack.
The announcement highlights how competition in artificial intelligence is shifting. The focus is no longer just about building bigger AI models. Companies are now racing to create the hardware, software, and cloud systems needed to support increasingly autonomous AI applications.
A Processor Designed for the Next Generation of AI
The newly announced Zhenwu M890 chip was developed by Alibaba’s semiconductor division, T-Head. According to the company, the processor offers roughly three times the performance of the earlier Zhenwu 810E model.
However, the most important part of the launch is not simply the speed increase. The M890 has been engineered specifically for AI agents — software systems capable of handling long tasks, maintaining context for extended periods, coordinating with multiple models, and carrying out complex operations with minimal human input.
These workloads place very different demands on hardware compared to traditional AI inference systems. AI agents require stronger memory performance, faster communication between computing components, and improved efficiency for continuous operation.
By designing a chip around these requirements, Alibaba is effectively preparing for what it believes will become the next major phase of enterprise AI adoption.
Alibaba Is Thinking Beyond Current AI Demand
Many AI chips today are optimized mainly for standard inference tasks, such as generating responses, analyzing images, or processing simple commands. Alibaba’s approach suggests the company expects AI systems to evolve into more independent digital assistants capable of managing multi-step workflows across industries.
That change could significantly alter future computing requirements. AI agents may need to run continuously for long periods, interact with multiple systems simultaneously, and make decisions in real time.
The M890 appears to be designed with that future in mind rather than focusing only on today’s mainstream use cases.
A Long-Term Semiconductor Roadmap
Alibaba also used the announcement to unveil a multi-year development plan for its AI chips.
The company stated that the M890 will eventually be followed by the V900 processor in 2027, which is expected to deliver another major performance improvement. After that, Alibaba plans to introduce the J900 chip in 2028.
This structured release cycle shows that Alibaba is building a sustained semiconductor strategy rather than treating chip development as a temporary response to current market conditions.
The roadmap mirrors the long-term upgrade cycles used by leading global chipmakers, particularly companies dominating the AI accelerator market.
China’s Tech Industry Is Reducing Dependence on Foreign Chips
Alibaba’s strategy also reflects a broader trend across China’s technology sector. Major Chinese companies increasingly view dependence on foreign semiconductor suppliers as a long-term business risk.
In recent years, export restrictions and supply chain concerns have encouraged domestic firms to invest more aggressively in self-developed chips and infrastructure.
Alibaba is not alone in this effort. Several large Chinese technology companies have expanded their own semiconductor initiatives, focusing on AI accelerators, cloud computing hardware, and advanced data processing systems.
Rather than relying entirely on imported hardware, these companies are building internal capabilities designed to support future AI growth regardless of external market pressures.
Massive Investment in AI Infrastructure
Alibaba’s semiconductor expansion is backed by significant financial investment.
The company previously committed more than 380 billion yuan toward AI and cloud infrastructure development over a three-year period. This represents one of the largest technology investment plans in Alibaba’s history.
The M890 chip and the broader silicon roadmap are direct outcomes of that long-term spending strategy.
By investing simultaneously in hardware, cloud systems, and AI software, Alibaba is attempting to create stronger control over its entire technology ecosystem.
Real-World Adoption Already Underway
T-Head revealed that more than 560,000 Zhenwu chips have already been shipped across multiple industries. According to the company, over 400 external customers are using the hardware in sectors including automotive and financial services.
This indicates that Alibaba’s AI hardware business has already moved beyond the experimental phase. The company is now gathering real-world deployment experience from commercial customers at scale.
Such adoption also gives Alibaba valuable operational data that can help improve future processors and AI systems.
AI Chips Integrated Into Alibaba Cloud
The new M890 chip will be available through Alibaba Cloud’s domestic AI platform known as Bailian.
Alibaba also introduced a server system called Panjiu AL128, which combines 128 M890 accelerators into a single rack configuration. This setup is designed to support large-scale enterprise AI workloads requiring substantial computing power.
By integrating its chips directly into its cloud services, Alibaba is creating tighter coordination between hardware and software performance.
This type of vertical integration has become increasingly important in the global AI industry, where efficiency and optimization can provide major competitive advantages.
Qwen 3.7-Max Expands the Software Side
Alongside the chip launch, Alibaba also introduced Qwen 3.7-Max, the newest version of its large language model.
The company says the model is optimized for advanced coding tasks and long-duration AI agent operations. Alibaba claims the system can operate continuously for extended periods without noticeable performance decline.
While such claims will likely be evaluated further over time, the launch clearly demonstrates Alibaba’s strategy of aligning its hardware and software development efforts.
The company is not simply releasing isolated AI products. It is building interconnected systems where chips, cloud services, and AI models are designed to work together.
The AI Competition Is Becoming More Strategic
Alibaba’s latest moves show how the AI race is rapidly evolving beyond chatbots and standalone models.
The industry is increasingly focused on building complete ecosystems that combine computing hardware, AI models, cloud infrastructure, and enterprise deployment tools into a single platform.
With a long-term chip roadmap already in place and large-scale infrastructure investment continuing, Alibaba appears committed to becoming a major player across every layer of the AI market.
As artificial intelligence systems become more advanced and autonomous, companies capable of controlling both hardware and software may hold a significant advantage in the years ahead.