Story

On air. The free wire runs 15 minutes behind live. See pricing

World, crypto, finance and tech news, scored for impact.

CryptoTokenPostKorean, translatedFiled 20 h ago

Alibaba trains Qwen on its own chip… V900 scheduled for mass production in Q1 2027

Original headline알리바바 자체 칩으로 큐원 훈련…V900 2027년 1분기 양산 예정

Open the sourceReported by TokenPost, tokenpost.kr, 20 November 2026, by 김하린

Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.

Alibaba confirmed during its September-quarter earnings call that it is training next-generation Qwen models on its in-house T-Head chips and serving inference for its Bailian platform on domestic silicon, extending homegrown hardware from training into production workloads. The company said its next AI chip, Zhenwu V900, will enter mass production in the first quarter of 2027, following its September 22 Apsara Conference unveiling, where Alibaba claimed three times the performance of the prior-generation M890 and the ability to scale to clusters of up to 500,000 chips. The disclosure matters because it signals Alibaba is reducing reliance on foreign accelerators across both model development and commercial AI services, a shift with implications for China's domestic semiconductor supply chain and for Nvidia's position in the market. What to watch next is whether Alibaba quantifies how much of its AI compute now runs on T-Head silicon, and whether the V900 timeline holds as Qwen training demands scale.

Muhabbit's AI wrote this summary in its own words from TokenPost's report and translated it from Korean. It can get details wrong, so check the source.

Named in this story

Alibaba (BABA), Qwen, Zhenwu V900, T-Head, Bailian

A linked name opens that asset's sentiment page.