Kingsoft Cloud Holdings Limited Q3 FY2025 earnings call

Strategic Positioning

• Firmly established strategic positioning in AI era, with technical and resource reserves for inference growth.

Revenue Growth

• Public cloud and enterprise cloud both saw y-o-y and sequential growth; public cloud up 49% y-o-y.

Intelligent Computing Cloud

• Gross billings of intelligent computing reached RMB 782,000,000, up ~122% y-o-y, accounting for 45% of public cloud revenue.

Ecosystem Revenue

• Revenue from Xiaomi and Kingsoft ecosystem was RMB 691,000,000, up 84% y-o-y, 28% of total revenue.

Enterprise Cloud Services

• Deeply explored verticals, built core competitiveness; examples in public services, healthcare, enterprise services.

Product and Technology

• Enhanced Intelligent Computing Cloud technology, launched model API service, upgraded online model services, and launched data annotation and dataset marketplace.

In the third quarter, Kingsoft Cloud's revenue reached RMB 2,480,000,000.00, with a year-over-year growth rate accelerating to 31%. Public cloud revenue was RMB 1,750,000,000.00, up 49% y-o-y. Intelligent computing cloud gross billings were RMB 782,000,000, up ~122% y-o-y, accounting for 45% of public cloud revenue. Revenue from the Xiaomi and Kingsoft ecosystem was RMB 691,000,000, up 84% y-o-y, making up 28% of total revenue. Adjusted gross profit for the quarter was RMB 393 million, up 28% y-o-y. Adjusted operating profit turned to profit at RMB 15.36 million, and adjusted net profit was RMB 28.73 million, a historical positive first-time profit.

Forward-Looking

• AI technology drives cloud computing revolution; will further invest in infrastructure, strengthen technology, enhance service stability, and provide high-value cloud services. Will continue to invest in infrastructure to support intelligent cloud business growth.

Q: Has there been any structural change in the demand of your ecosystem and external clients for the past quarter? And secondly, how does management see the margin trend in the coming quarters? And what's the expected mix of different computing resources acquisition models?

A: Zhou Tao mentioned AI revenue growth in Q3 due to partially delivered clusters and delayed revenue. Regarding margin, inference demand is expected to have higher margin than training. Li Yi stated EBITDA margin will remain above 20% but noted significant quarter-on-quarter improvement in Q3 was due to one-time other income.

Q: Could management share the outlook and guidance on the revenue outlook for next year? And beyond the Internet companies' post-model training and in-body intelligence scenarios that are already underway this year, which other industry and application scenarios are expected to have strong computing power demand that could drive the revenue growth next year? And with multiple providers in both China and the US increasing the proportion of server leasing in their computing resource mix, how does management view the current market dynamics for procurement versus leasing? And from a cost-effectiveness and profit margin perspective, how would the company allocate the resources between these two approaches?

A: Li Yi said budget process is underway; confident in AI business demand growth. Zhou Tao mentioned robotic companies in China and API services as growth drivers. Regarding procurement vs leasing, premium customers like Xiaomi tend to use CapEx model, while growth stage companies use leasing model. No top-down target for split, and margin expected to improve.

Q: Regarding the differences between AI training versus inferences. Could management share what is the pricing methodology between these two kinds of demand and what has been the part pricing trend over the past few months or year to date? And, in terms of the utilization rate of the chips of GPUs, pricing, and profitability, can you share more color on the gap between training and inferences?

A: Zhou Tao stated price is based on resource usage; inference services using platform have margin similar to training, while API talking services have better margin, but this business is in early stage.