• Home
• Business

Alibaba Group shares climbed approximately 3% in Hong Kong before extending gains to 5.1%—hitting a one-month high—after the Chinese tech conglomerate unveiled a next-generation artificial intelligence processor and outlined massive global data center expansion plans. Announced during the annual Apsara Conference hosted by Alibaba Cloud in Hangzhou, the roadmap signals an aggressive bid to scale foundational computing infrastructure amid intensifying global competition from Western hyperscalers and chipmakers.

Key Technology and Infrastructure Announcements

Alibaba outlined a comprehensive roadmap covering proprietary silicon, massive power capacity, and ultra-large frontier AI models:

Strategic PillarAnnouncement Details & Operational Scale

Zhenwu V900 AI ChipNext-generation AI processor delivering 3x the performance of its predecessor, the Zhenwu M890 (launched in May). Scheduled for mass production and commercial deployment in Q1 2027, optimized for large-scale cluster training. Existing Zhenwu chips are currently utilized by over 650 enterprise clients across automotive, finance, energy, and manufacturing.

Global Data Center ExpansionTargeting more than 20 Gigawatts (GW) of global data center computing capacity by 2032, scaling up cloud infrastructure to meet surging enterprise AI workloads.

Next-Gen Frontier ModelsConfirmed that Qwen 4 is currently in active training, with roadmaps established for Qwen 4.5 and Qwen 5. In parallel, Alibaba is preparing to train a massive frontier model scaling between 5 trillion and 10 trillion parameters—up to 4x the scale of its current flagship, Qwen 3.8 Max.

Global Infrastructure Race Context

Alibaba’s capital commitments reflect the broader escalation in global computing capacity deployments:

NVIDIA: Disclosed plans earlier this month with Australian partners to support up to 2 GW of dedicated AI infrastructure in Australia by 2027.

Meta: Announced plans in July to construct its first Canadian data center in Alberta—a 1 GW facility projected to cost approximately $9 billion with a three-year construction timeline.