China Cloud Computing Market 2025: Alibaba Huawei and Tencent Cloud
China's cloud computing market surpassed 600 billion RMB (83 billion USD) in 2025, driven by enterprise digital transformation, AI workload migration, and government cloud adoption. Alibaba Cloud leads with approximately 30% market share, followed by Huawei Cloud at 20% and Tencent Cloud at 15%. The market is undergoing a fundamental shift as AI workloads (LLM training, inference, and AI-native applications) become the primary growth driver.
TL;DR
China's cloud computing market reached 600 billion RMB in 2025. Alibaba Cloud maintained its lead at 30% market share, Huawei Cloud grew to 20%, and Tencent Cloud held 15%. AI-related cloud services grew 60% year-over-year, now accounting for 25% of total cloud revenue. GPU cloud spending exceeded 100 billion RMB as enterprises rushed to adopt LLM capabilities.
Key Insights
Total Market Size
China's cloud computing market reached 600 billion RMB, including IaaS, PaaS, and SaaS segments. Public cloud accounted for 65% while private and hybrid cloud solutions represented the remaining 35%.
Alibaba Cloud Revenue
Alibaba Cloud (Aliyun) generated approximately 180 billion RMB in revenue, maintaining its position as China's largest cloud provider. It serves over 4 million customers including 60% of China's listed companies.
GPU Cloud Spending
GPU cloud infrastructure spending exceeded 100 billion RMB, driven by demand for LLM training and inference. Chinese cloud providers invested heavily in GPU clusters using NVIDIA A100/H100 chips and domestic alternatives like Huawei Ascend.
AI Cloud Services Growth
AI-related cloud services grew 60% year-over-year, now accounting for 25% of total cloud revenue. Key services include model-as-a-service, AI inference APIs, and vector database offerings.
Side-by-Side Comparison
| Provider | Market Share | Revenue | Key Strength | AI Focus |
|---|---|---|---|---|
| Alibaba Cloud | 30% | 180B RMB | E-commerce + enterprise | Tongyi Qianwen MaaS |
| Huawei Cloud | 20% | 120B RMB | Government + telco | Pangu models + Ascend |
| Tencent Cloud | 15% | 90B RMB | Gaming + social | Hunyuan MaaS |
| Baidu AI Cloud | 8% | 48B RMB | AI + search | ERNIE ecosystem |
| AWS China | 5% | 30B RMB | Multinational | Global consistency |
| Others | 22% | 132B RMB | Various | Various |
Frequently Asked Questions
China's cloud computing market differs from the US in several key ways: market structure (China's market is more fragmented with domestic players dominating, while the US market is concentrated around AWS, Azure, and GCP); regulatory environment (data localization laws require Chinese citizen data to be stored domestically, creating natural barriers for foreign providers); AI integration (Chinese cloud providers have been faster to integrate AI capabilities due to government encouragement and the booming domestic AI market); pricing (Chinese cloud services are generally 30-50% cheaper than equivalent US services due to intense domestic competition); government adoption (Chinese government and SOE cloud spending accounts for 25% of the market, compared to approximately 15% in the US); and GPU availability (US export restrictions on advanced NVIDIA chips to China have created challenges, pushing Chinese providers to invest in domestic GPU alternatives like Huawei Ascend). The Chinese market is also growing faster (25% vs 15% in the US) as enterprise digital transformation in China still has significant runway compared to the more mature US market.
US export restrictions on advanced AI chips (NVIDIA A100, H100, and later H200) have had significant but manageable impacts on Chinese cloud providers. Key effects include: GPU procurement costs have increased 30-50% as providers must use less efficient chips (NVIDIA A800, H800 which are downgraded export versions) or domestic alternatives; Huawei's Ascend 910B has emerged as the primary domestic GPU alternative, with Huawei Cloud now offering competitive AI training and inference services using Ascend chips; the restrictions have accelerated domestic GPU development, with companies like Biren, Moore Threads, and Iluvatar CoreX making progress on competitive chips; Chinese cloud providers have developed sophisticated model optimization techniques to maintain performance on less powerful hardware, including model compression, quantization, and efficient training algorithms; and the restrictions have paradoxically benefited Huawei Cloud, which has its own chip supply chain. Most Chinese cloud providers report that while the restrictions create additional costs and complexity, they have not fundamentally limited their ability to offer competitive AI cloud services, particularly for inference workloads which are less GPU-intensive than training.