Alibaba Launches Zhenwu M890 AI Chip to Challenge Nvidia
Alibaba has launched the Zhenwu M890, an AI chip designed to reduce China’s reliance on Nvidia hardware amid ongoing US export controls. Alibaba’s semiconductor unit, T-Head, says the M890 delivers three times the performance of the previous Zhenwu 810E and is built for training and operating agentic AI systems.
The Zhenwu M890 includes 144GB of on-chip memory and 800GB/s of inter-chip bandwidth. It is already being deployed in Alibaba Cloud’s Panjiu AL128 supernode server. Using the ICN Switch 1.0, the system can connect up to 64 accelerators in congestion-free clusters.
T-Head has shipped more than 560,000 Zhenwu chips to over 400 customers, including China Telecom and FAW Group. External clients now account for about 60% of its production capacity, indicating growing demand for domestic AI hardware in China’s data-centre and technology sectors.
Alibaba plans to launch the Zhenwu V900 in the third quarter of 2027, with performance estimated at three times that of the M890. The J900 is scheduled for 2028. For AI and semiconductor investors, the rollout highlights China’s push for supply-chain independence and could influence sentiment toward data-centre, chipmaking and cloud-computing companies. The immediate cryptocurrency impact is limited because the announcement contains no direct blockchain or token-related development.
Neutral
The expected cryptocurrency-market impact is neutral. Alibaba’s Zhenwu M890 is significant for AI hardware, cloud computing and China’s semiconductor supply chain, but the announcement does not involve a blockchain network, crypto asset, mining technology or digital-asset regulation.
In the short term, traders may see limited spillover into AI-linked tokens or broader technology sentiment. Similar announcements from major technology companies can briefly lift AI-related assets when investors interpret them as evidence of stronger infrastructure demand. However, such moves are usually sentiment-driven and tend to fade without direct effects on crypto revenues, token utility or network activity.
Over the longer term, faster development of domestic AI chips could support demand for data-centre power, connectivity and computing infrastructure. That may indirectly benefit crypto projects focused on decentralised computing or AI. It could also create competitive pressure for US chipmakers and affect technology-sector risk appetite. Nevertheless, there is no clear catalyst for Bitcoin, Ethereum or major altcoins. Traders should therefore monitor follow-up evidence, such as Alibaba’s production volumes, customer adoption, AI-capex trends and any partnerships with blockchain or decentralised-computing projects, rather than treating the launch as a standalone crypto signal.