Alibaba has been making significant strides in AI. The company announced a high-performance new AI chip.
Additionally, it also revealed plans to release a new giant model. The new Alibaba AI chip is designed to reduce the cost of computing. That is, Alibaba desires scale and efficiency jointly.
Announced at one of the company’s flagship events, Alibaba’s annual Apsara Conference was held in Hangzhou. The announcement was on 22nd September 2026, with Chief Executive Eddie Wu leading the presentation.
The key announcements were:
- New chip: The Zhenwu V900, called China’s most powerful AI chip
- Performance: Triple the power of its predecessor
- Release: Commercial launch planned for early 2027
- New model: An AI model with 5 to 10 trillion parameters
- Cloud goal: Data centre capacity above 20 gigawatts by 2032
| Item | Detail |
|---|---|
| Chip name | Zhenwu V900 |
| Chip claim | Most powerful AI chip in China |
| Model size | 5 trillion to 10 trillion parameters |
| Announced by | CEO Eddie Wu |
| Event | Apsara Conference, Hangzhou |
| Cloud target | Over 20 GW capacity by 2032 |
Why the Chip and Model matter?
Alibaba’s plan ties hardware and software together. Each part supports the other closely. The sections below explain the cost logic and the politics. Both help explain the timing of this launch.
1. Cutting AI Costs
The cost angle sits at the heart of the strategy. Powerful in-house chips can lower training expenses. They reduce dependence on costly imported hardware. That matters as AI models grow ever larger. In other words, controlling the chip controls the cost.
Scale on this level demands enormous computing power. Alibaba is expanding its cloud infrastructure to match. It wants Alibaba Cloud capacity above 20 gigawatts by 2032. Therefore, the chip and the cloud plans align tightly.
2. The Geopolitical Angle
The timing carries clear geopolitical weight. The announcements came amid US-China tech tension. Chinese and US leaders prepared to meet in Washington. AI competition featured high on that agenda. Consequently, domestic chips have national importance. In addition, they support China’s push for self-reliance.
Alibaba is not acting alone in this effort. Huawei and other firms build rival chips too. Together they form a growing domestic ecosystem. That ecosystem could reduce reliance on Nvidia.
Strategic Implications for B2B Buyers
The move has broad implications for business. Cheaper, more powerful AI could reach more firms. Chinese companies may gain access to advanced tools. That could boost productivity across many sectors.
Enterprises increasingly depend on cloud AI for operations. Reliable and affordable compute underpins those efforts. A global B2B marketplace helps firms find technology partners. Such platforms connect buyers with suppliers across regions.
The chip launch fits Alibaba’s wider AI spending drive. B2BInside covered how Alibaba grew revenue while AI spending cut profits. The whole sector is chasing cheaper AI too. Our report on Anthropic's Opus 5 at half the price shows the trend. In other words, cost control now shapes AI strategy.
Frequently Asked Questions (FAQs)
What is Alibaba's new AI chip?
It is called the Zhenwu V900. Alibaba calls it China’s most powerful AI chip. It offers triple the performance of its predecessor.
How big is Alibaba's planned AI model?
The model will carry 5 to 10 trillion parameters. That would rank among the largest models anywhere. Bigger models can handle more complex tasks.
Who announced Alibaba's AI plans?
Chief Executive Eddie Wu announced them. He spoke at the Apsara Conference in Hangzhou. The news broke on September 22, 2026.
How does the chip cut AI costs?
In-house chips reduce reliance on costly imports. They can lower the expense of training big models. That helps Alibaba control its overall AI costs.
What is Alibaba Cloud's data centre goal?
Alibaba wants capacity above 20 gigawatts by 2032. That scale supports its huge new AI models. It also strengthens Alibaba Cloud’s competitiveness.
