Alibaba raises prices for its T-Head AI computing chips, including the Zhenwu 810E, by 5% to 34%, and its Cloud Parallel File Storage by 30%, after demand soars
Alibaba Group Holding Ltd. is raising prices for its AI computing and storage products by as much as 34% …
Context & Ripple Effects
Alibaba’s pricing turn reverses its earlier cloud-cost push: it cut prices on more than 100 core cloud products in 2024, then made another broad reduction across cloud offerings. The shift matters because it reaches both AI compute hardware and the storage layer used alongside it.
The move was not isolated: Baidu subsequently raised prices for AI-computing services and parallel file storage, while later coverage recorded higher pricing for an advanced AI model. Together, these developments suggest demand is changing the competitive calculus from discounting toward capacity monetization.
First-order effects
- Alibaba customers buying T-Head AI chips, including Zhenwu 810E, face higher acquisition costs, while users of Cloud Parallel File Storage face a 30% increase.
- Alibaba can capture more revenue per unit of constrained AI compute and associated storage rather than using those products primarily as cloud-price competition tools.
Second-order effects
- Customers building or expanding AI workloads may reassess chip, compute-service, and storage configurations as the combined infrastructure bill rises.
- Peer providers face stronger cover to reprice capacity: Baidu’s parallel move on AI-computing and storage pricing indicates the pressure is extending beyond one supplier.
Third-order effects
- If such increases persist, AI infrastructure competition may be shaped less by headline cloud discounts and more by the availability and monetization of compute and high-performance data services.
- The pattern would strengthen the link between AI workload demand and pricing across adjacent infrastructure layers, from chips to storage and hosted compute; its durability depends on whether capacity catches up with demand.
The trend: AI infrastructure providers are moving from cloud-price competition toward pricing scarce compute and the storage services that support AI workloads.