Alibaba: Unveils Zhenwu V900 AI Chip, China's Most Powerful

• Alibaba unveiled its new Zhenwu V900 AI chip, which it claims is China's most powerful, delivering three times the performance of its predecessor.
• The company plans to train a new AI model at a scale of five to 10 trillion parameters and expand its data centers to over 20 gigawatts of computing by 2032.
• This announcement comes amidst a broader Chinese push for AI self-reliance, with other firms like Moonshot and Huawei also making significant advancements in chips and models.
• Analysts suggest China's AI progress provides leverage in upcoming U.S.-China talks, where AI, trade, and tariffs are expected to be key discussion points.
• The developments highlight the ongoing US-China technology rivalry, intensified by U.S. export restrictions on advanced AI chips and chipmaking equipment.

The timing of Alibaba's significant AI announcements, made just days before a scheduled meeting between Chinese and U.S. leaders, highlights the intensifying US-China technology rivalry.

Alibaba, the Chinese technology giant, recently announced significant advancements in its artificial intelligence capabilities, including the introduction of a new AI chip and plans for more powerful AI models. The company revealed its Zhenwu V900 AI chip, which it claims is currently the most powerful AI chip developed in China. This new chip reportedly delivers three times the performance of Alibaba's previous generation Zhenwu M890 chip, playing a crucial role in powering frontier AI model training and the complex calculations (inference) required for AI responses.

These Zhenwu chips are integral to Alibaba's data centers, providing essential computing power for both the company's internal operations and its cloud clients. Alongside hardware innovations, Alibaba outlined its strategy to develop a new AI model with an impressive scale of five to 10 trillion parameters, a metric indicating an AI's learning capacity. This ambitious target aims to bring Alibaba's AI models closer to the most advanced offerings from the United States. For context, its current most powerful Qwen AI series model, the Qwen3.8-Max, operates with 2.4 trillion parameters.

Further demonstrating its commitment to AI expansion, Alibaba also highlighted plans to significantly expand its data center infrastructure to meet what it describes as an exponentially rising demand for AI computing. The company projects achieving over 20 gigawatts of computing capacity by 2032. This target surpasses the approximately 1.4 gigawatts of AI computing capacity held by SpaceX as of mid-this year, though SpaceX aims to reach over 10 gigawatts by 2027. Alibaba CEO Eddie Wu emphasized the company's dedication to mobilizing every available resource to satisfy customer demand for AI solutions.

Alibaba's announcements are part of a broader trend of rapid AI development within China, where other domestic players are also making significant strides. For instance, Chinese AI company Moonshot stated that its Kimi K3, launched in July, holds the distinction of being the world's largest open model, featuring 2.8 trillion parameters. This competitive environment underscores China's push for self-reliance in advanced technologies, particularly as U.S.-led export restrictions continue to limit its access to some of the world's most cutting-edge technologies, including advanced AI chips and chipmaking machinery.

The landscape of China's AI chip competition is further intensified by the efforts of other technology giants. Chinese technology firm Huawei, for example, recently unveiled its own new chip technologies, signaling its intent to challenge global leaders such as Nvidia. While frontier AI model training in China has historically relied heavily on Nvidia chips, Chinese-designed chips are steadily gaining market share, largely due to the U.S. government's restrictions preventing Nvidia from selling some of its most powerful AI chips to China.

Analysts note that this domestic push is crucial for China's technological resilience. Neil Shah, Vice President of Counterpoint Research, observed that leveraging additional computing power to compensate for chip limitations helps China maintain a strong local presence. Parv Sharma, a senior analyst at the same firm, added that the extent to which the AI and chipmaking gap between China and the U.S. narrows will depend not only on China's chip design capabilities but also on its ability to advance the foundries that manufacture these sophisticated components.

The timing of Alibaba's significant AI announcements, made just days before a scheduled meeting between Chinese and U.S. leaders, highlights the intensifying US-China technology rivalry. Chinese leader Xi Jinping was set to arrive in Washington for a state visit and meeting with U.S. President Donald Trump, where discussions on AI, trade, and tariffs were anticipated to be central themes. This development underscores the strategic importance of AI leadership in global geopolitics.

Some analysts believe that recent Chinese advancements in AI provide Beijing with increased leverage in these high-stakes discussions with Washington. Concurrently, U.S. AI leaders, including Anthropic CEO Dario Amodei, have voiced concerns that China's rapid AI development poses a threat to the United States, advocating for an overall slowdown in the technology's progression. This divergence in perspectives sets the stage for potential policy shifts and increased scrutiny over AI-related trade and technology transfers.

As China continues its drive for technological self-reliance, particularly in response to existing U.S. export restrictions, the global availability and adoption of Chinese AI models are also expanding. Open Chinese models, often more affordable than advanced closed models from leading U.S. AI laboratories, have begun making inroads internationally, including within the United States, further complicating the competitive landscape and raising questions about future regulatory frameworks.

This development highlights the intensifying US-China technology rivalry, signaling potential for new or tightened US export controls and trade restrictions on AI-related technologies. Lawyers and compliance officers should monitor upcoming US-China policy discussions and legislative actions that could impact US companies operating in or with the AI and semiconductor sectors.

Source: Original reporting via Associated Press