Published • loading... • Updated
Alibaba deepens AI push with new chip, bigger model; shares jump 5%
The new chip is designed to triple performance and scale to clusters of 500,000 units, while Alibaba targets a 5 trillion-parameter Qwen model.
- On Tuesday, Alibaba Group Chief Executive Eddie Wu announced a strategy targeting 20 gigawatts of Alibaba Cloud data center capacity by 2032, citing "exponentially rising demand for AI" at the Apsara Conference in Hangzhou.
- The company plans to scale upcoming Qwen series models to between 5 trillion and 10 trillion parameters, significantly exceeding the 2.4 trillion parameters of the current flagship Qwen 3.8 Max.
- Alibaba also introduced the Zhenwu V900, a next-generation AI chip from its T-Head unit that Eddie Wu claimed delivers three times the performance of the M890 supernode and supports up to 500,000 cards.
- Scheduled for mass production in the first quarter of 2027, the chip arrives as Chinese firms race to develop domestic alternatives to Nvidia processors amid ongoing U.S. export restrictions.
- The Qwen team continues research into recursive self-improvement to complete "more complex, longer-horizon tasks" and advance toward artificial superintelligence, or ASI, while Alibaba expects "significant growth" in annual AI chip shipments.
Insights by Ground AI
Podcasts & Opinions
167 Articles
167 Articles
INTELLINEWSReposted by 
Courrier international
Courrier international
The Chinese firm Alibaba presented new artificial intelligence chip technologies and plans for more powerful models on Tuesday, including what it claimed to be China's most powerful AI chip, a few days from a meeting between Chinese and U.S. leaders in which competition for leading AI technology is expected to be a central issue.
·El Segundo, United States
Read Full Article
Alibaba introduced its new Zhenwu V900 accelerator, which triples the performance of the previous chip.
·Mexico
Read Full ArticleCoverage Details
Total News Sources167
Leaning Left25Leaning Right26Center37Last UpdatedBias Distribution42% Center
Bias Distribution
- 42% of the sources are Center
42% Center
L 28%
C 42%
R 30%
Factuality
To view factuality data please Upgrade to Premium
