Wall Street Commentary on Alibaba Cloud Computing Conference: 20GW Computing Power Target Achieved, Cloud Revenue Ceiling Raised Again

Wall Street Commentary on Alibaba Cloud Computing Conference: 20GW Computing Power Target Achieved, Cloud Revenue Ceiling Raised Again

Following the conclusion of the Yunqi Computing Conference, major Wall Street investment banks quickly released research reports, providing a systematic interpretation of Alibaba's strategic layout from three dimensions: capital expenditure path, cloud revenue ceiling, and AI full-stack competitiveness. Goldman Sachs, Citigroup, and UBS all maintained their buy ratings, but the three institutions gave different calculations on the cloud revenue scale implied by the 20GW computing power target, reflecting the core controversy in the market regarding the pace of Alibaba's AI monetization.

At the conference, Alibaba CEO Wu Yongming announced a goal to expand the total power consumption of data centers to 20GW by 2032, approximately ten times that of 2022. This statement was regarded by three investment banks as the most valuable incremental information at the conference, directly triggering a recalculation of the long-term ceiling for cloud revenue.

Citigroup believes that the 20GW target corresponds to approximately $160 billion in external cloud revenue in fiscal year 2033, an increase of approximately $60 billion from its previous forecast of $100 billion in fiscal year 2031; UBS, based on different assumptions, estimates the external cloud revenue corresponding to the same target in 2032 to be approximately $170 billion, both implying a compound annual growth rate of approximately 40%.

All three institutions maintained their buy ratings, but their target prices differed significantly. Goldman Sachs gave Alibaba's US-listed shares a target price of $177, Citigroup gave $190, while UBS was the most aggressive, with a target price of $206, representing approximately 78% upside potential from the current share price.

20GW target: Cloud revenue ceiling rises significantly, but there are disagreements on capital expenditure paths.

The 20GW expansion target is a common focus of the three investment bank reports, but there are differences in their judgments on the scale of capital expenditure and financing path.

According to Goldman Sachs' calculations, expanding from the current 5-6GW to 20GW by 2032 implies an annual increase of 1-2GW in the coming years, corresponding to Alibaba's annual capital expenditure of approximately RMB 200-300 billion. Goldman Sachs projects Alibaba's capital expenditure for FY27-FY29 to be RMB 209 billion, RMB 243 billion, and RMB 260 billion, respectively. Goldman Sachs emphasizes that this scale is within a manageable range and will be achieved through a combination of three models: self-built heavy assets, light asset partnerships, and leasing. Funding sources include stable EBITA from the e-commerce business, profit growth from the cloud business, cash reserves, and future financing.

UBS expects Alibaba's FY27 capital expenditure to reach RMB 230 billion, noting that the expansion plan will rely on three paths: self-built capacity (supported by approximately RMB 190 billion in cash flow from Taotian Group), external leasing (included in operating costs), and partnerships.

Citigroup pointed out that Alibaba has not yet updated its previous capital expenditure guidance of RMB 380 billion, but expects capital expenditure to remain high in the coming years, with some infrastructure capacity to be supplemented through operational cooperation with industry partners.

Self-developed chip "PingtouGe": Cost advantage may become a key differentiator.

Goldman Sachs gave significant weight to Alibaba's self-developed chip "PingtouGe" in its report, considering it key to understanding the economics of the 20GW target. Goldman Sachs pointed out that PingtouGe's self-developed chip approach has a significant USD/GW cost advantage compared to similar domestic and international solutions, and expects PingtouGe to contribute about half of Alibaba Cloud's computing power in the medium term, compared to only about 10% currently.

The new-generation AI chip, Zhenwu V900, launched at the conference boasts performance three times that of the current M890, supports single clusters of up to 500,000 cards, and is planned to eventually expand to clusters of millions of cards. Goldman Sachs believes that shipments of the "PingtouGe" chip will increase significantly in the coming years, which will directly reduce the cost per unit of computing power and improve the capital expenditure conversion efficiency of cloud businesses.

UBS also mentioned in its report that Alibaba plans to launch its self-developed Yitian 720/730 CPU in 2027, which will work in synergy with the V900 to further improve the cost-effectiveness of computing power.

Thousand Questions Model: ARR Accelerated Realization, Parameter Scale Leaps to 5-10T

At the model level, all three investment banks have focused on the commercialization progress of Qianwen's MaaS platform (Bailian). According to Goldman Sachs, citing company data, Qianwen AI MaaS platform's annualized recurring revenue (ARR) reached RMB 20 billion at the end of August, continuing to accelerate from RMB 8 billion in mid-May and RMB 16 billion in mid-August. The company maintains its target of reaching RMB 30 billion by the end of the year.

Regarding its model roadmap, Alibaba has clarified that the Thousand Questions 4 series is about to be released, and Thousand Questions 4.5 and Thousand Questions 5 will evolve towards a scale of 5 to 10 trillion parameters, while continuously strengthening multimodal capabilities and promoting recursive self-improvement (RSI). Goldman Sachs pointed out that Thousand Questions 3.8 Max's user revenue increased 8.5 times in two months, and token consumption surged 12 times, with efficiency improvements and accelerated commercialization progressing simultaneously.

Goldman Sachs maintained its higher-than-market consensus forecast for Alibaba Cloud's revenue growth, projecting year-over-year growth of 53%, 55%, and 55% for the September, December, and March 2026 quarters, respectively.

In its report, UBS systematically reviewed management's assessment of AI cloud demand and identified six key structural drivers:

First, the modalities are expanding from text to images, videos, audio, and 3D world models, broadening the addressable scenarios and increasing computing power requirements;

Second, inference needs surpass training needs, expanding the user base from model developers to end users and significantly increasing the overall addressable market.

Third, AI has a multiplier effect on traditional cloud services, driving cross-selling of storage, network, security and other products;

Fourth, demand is rapidly penetrating into industries beyond the internet, such as manufacturing, finance, autonomous driving, robotics, and agriculture.

Fifth, mobile and PC applications will be restructured around AI, driving a new round of cloud migration;

Sixth, the growth rate of overseas business continues to outpace that of domestic business, and Alibaba Cloud is actively expanding its data centers in South America, the Middle East and Europe.

Citigroup also pointed out that the main constraint Alibaba Cloud currently faces is still supply rather than demand. The tight supply of AI computing power is the background for the acceleration of cloud revenue, which is also the direct driving logic for the 20GW expansion target.

Valuation and Risks: Target Price Divergence Reflects Uncertainty in Realization Pace

The differences in the target prices of the three institutions reflect, to some extent, the different assumptions that each party has about the pace of return on investment in Alibaba AI.

Goldman Sachs used the SOTP valuation method to give Alibaba a 12-month target price of US$177 for its US-listed shares and HK$172 for its Hong Kong-listed shares, and expects earnings per share to grow by 58% and 27% year-on-year in FY27 and FY28, respectively.

Citigroup and UBS gave target prices of $190 and $206 respectively. The latter believes that as the return on AI investment improves, the market will gradually raise earnings forecasts and refocus on the value of Alibaba's AI assets.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.