Global Cloud Market Q3 2025: AWS Share Declines to 29%, AI Drives Record Market Growth

Introduction: Cloud Market Enters AI-Driven Boom Period

In the third quarter of 2025, the global cloud infrastructure services market reached a landmark moment. According to the latest data from Synergy Research Group, the market size in that quarter reached $107 billion, up 28% year-over-year and up $7.6 billion quarter-over-quarter, setting a record for the largest quarterly growth ever. This data reveals a key signal: AI has transformed from a technical concept into the core engine of cloud market growth. At the same time, the market share landscape of leaders is quietly shifting—AWS fell from 31% in the same period last year to 29%, declining for several consecutive years; Microsoft and Google remained stable at 20% and 13% respectively, demonstrating resilience. This industrial transformation triggered by AI is redefining the evolution direction of enterprise IT architecture.

Event Background: Three Giants Dominate, but AWS Loses Momentum

According to the Q3 2025 global cloud market share data released by Synergy Research Group, AWS, Microsoft, and Google Cloud together hold 62% of the market share. Among them, AWS continues to lead with 29% share, but has shown a year-over-year downward trend from 31% in Q3 2024 and 34% in Q3 2022. Microsoft ranks second firmly with 20% share, with its intelligent cloud division generating quarterly revenue of $30.9 billion, up 28% year-over-year, with an annualized run rate of approximately $123 billion. Google Cloud ranks third with 13% share, with revenue of $15.2 billion, up 34% year-over-year, with an annualized run rate of approximately $61 billion.

Among Chinese cloud vendors, Alibaba Cloud ranks fourth with 4% share, with revenue of $4.7 billion, up 26% year-over-year; Oracle ranks fifth with 3% share, with cloud revenue of $7.2 billion, up 28% year-over-year. In addition, Salesforce, IBM, Tencent Cloud, and Huawei Cloud each account for about 2%, while multiple vendors including Akamai, Baidu, China Telecom, China Unicom, CoreWeave, Databricks, Fujitsu, NTT, OpenAI, Snowflake, and SAP each account for about 1%.

John Dinsdale, chief analyst at Synergy Research, pointed out: "The market is going from strength to strength, with record-breaking numbers. The sequential increase of $7.6 billion is the largest ever." He emphasized that AI has penetrated every aspect of these numbers, driving the rapid expansion of AI-specific services while also boosting growth in other cloud services. GPUaaS (GPU as a Service) revenue has grown at an annual rate of over 200%, serving as the most direct example of the AI-driven cloud market.

Technical Analysis: How AI Reshapes Cloud Infrastructure DemandAI, especially generative AI, has fundamentally different demands on cloud infrastructure than traditional enterprise workloads. Training large language models requires massive GPU clusters, which place extremely high requirements on computing density, network bandwidth, storage speed, and power supply. Traditional cloud data centers are designed around CPUs and virtualization, while data centers in the AI era need tight collaboration among accelerator chips such as GPUs and TPUs, prompting cloud providers to accelerate the construction of AI-specific infrastructure.

GPUaaS is a typical product of this trend. It allows enterprises to rent GPU computing power as a service, eliminating the need to build expensive AI clusters of their own, thereby lowering the barrier to adopting AI technology. According to Synergy data, GPUaaS revenue is growing by more than 200% annually, reflecting strong enterprise demand for AI training and inference computing power. In addition, AI has also driven growth in adjacent markets such as cloud databases, data analytics, and security services, because these services need to be deeply integrated with AI workloads.

For enterprises, this means the value proposition of cloud services is shifting. In the past, enterprises moved to the cloud mainly to reduce IT costs and improve elasticity; now, cloud platforms have become a key channel for enterprises to acquire AI capabilities and achieve data-driven transformation. The capabilities of AI infrastructure are becoming one of the core considerations when enterprises choose cloud providers.

Enterprise Impact Analysis: The Dual Considerations of Cost and Strategy

The explosion of AI cloud services has brought new complexity and opportunities to enterprise IT decision-making.

Cost impact: The price of AI computing power is far higher than that of traditional computing resources. If enterprises use GPUaaS or AI platform services heavily, their cloud bills (OPEX) may grow exponentially. On the other hand, the CAPEX (capital expenditure) of building AI infrastructure in-house is equally high, including the procurement of GPU servers and the renovation of data center power and cooling systems. For most enterprises, renting AI services from the cloud may be more economical in the short term, but in the long run, frequent AI training tasks may prompt large enterprises to consider a hybrid strategy—placing core training workloads in their own facilities and outsourcing elastic demand to the cloud.

Deployment impact: AI workloads place higher demands on the physical location of data centers. Latency-sensitive AI applications (such as autonomous driving and real-time inference) need edge nodes close to data sources, while training tasks are concentrated in data-center-dense regions. This requires enterprises to strike a more refined balance between multi-cloud and edge computing. Cloud providers are expanding data centers globally to meet AI's demand for geographic distribution of computing power.Operational Impact: The operational complexity of AI infrastructure is far higher than that of traditional cloud services. GPU cluster failure rates, energy consumption management, and network optimization all require specialized teams. Many enterprises lack such in-house capabilities, so they tend to prefer managed AI cloud services, shifting the operational burden to cloud providers. This drives cloud providers to offer higher-level AI platforms, such as pre-trained models, automatic hyperparameter tuning, data pipelines, etc., helping enterprises reduce the complexity of AI operations.

Security and Compliance: AI training involves large amounts of sensitive data, and data sovereignty and privacy protection have become issues that enterprises must consider when moving to the cloud. As a result, demand for Sovereign Cloud is rising; enterprises want cloud providers to offer data residency, access control, and compliance certifications. At the same time, security risks inherent to AI models themselves (such as model leakage and adversarial attacks) also require cloud platforms to provide specialized security tools.

Taken together, enterprises should reassess their cloud strategies, no longer basing decisions solely on price or features, but instead considering AI capabilities, data governance, ecosystem integration, and long-term costs. For enterprises planning large-scale AI adoption, now is the time to establish deeper partnerships with cloud providers to optimize compute procurement and architecture design.

Market Competition Analysis: AWS's Share Erosion and Opportunities for Challengers

AWS's share fell from 34% in 2022 to 29%, which does not mean its business is stagnating, but rather that the new demand brought by AI within market growth is being captured by other vendors. Microsoft, leveraging the deep integration of Azure and OpenAI, has taken a first-mover advantage in the AI cloud space. Its intelligent cloud revenue grew 28% year-over-year, and the quarterly revenue gap with AWS is only about $3 billion, with faster growth. Google is also actively investing in AI infrastructure and platform services, with a contract backlog reaching $10.6 billion, and its growth momentum deserves attention.

Alibaba Cloud and Oracle also performed commendably. Alibaba Cloud dominates the Chinese market while maintaining stable share through international expansion. Oracle, through partnerships with AI chip vendors and differentiated positioning in cloud databases, has established a foothold in the cloud infrastructure market. Although its share is only 3%, its cloud revenue growth rate reached 28%, indicating strong pricing power in specific verticals such as finance and telecommunications.

The overall performance of Chinese cloud vendors is affected by slowing domestic market growth and international geopolitics, but Tencent Cloud and Huawei Cloud still maintain about 2% global share each. As Chinese enterprises advance their globalization, these cloud vendors may become the preferred cloud platforms for Chinese companies going overseas.

For AWS, the market share decline is a warning signal. Although it remains the undisputed leader, in the AI era, the technology and ecosystem gap with competitors is narrowing. AWS needs to accelerate the launch of more competitive AI services and address GPU supply and energy consumption issues; otherwise, it may face the risk of further losses. Microsoft and Google, on the other hand, need to be wary of an AWS counterattack while maintaining their own partnership advantages in the AI field.For emerging cloud vendors such as CoreWeave and Databricks, AI brings opportunities for differentiated competition. These vendors focus on AI infrastructure or data services and can respond to specific needs more quickly, but they still find it difficult to shake the market leaders in terms of scale and ecosystem. Competition in the cloud market will become more diversified and fragmented, but the leader effect remains evident.

Industry Trend Watch: AI-Native Cloud Becomes a Definitive Direction

The Q3 data confirms that the cloud market is fully entering the AI-native era. The so-called AI-native cloud means that everything from infrastructure to platform services is designed around AI workloads, rather than simply adding AI features on top of traditional cloud. This is reflected in the following trends:

• AI infrastructure spending continues to grow: GPUaaS revenue has grown by more than 200% year over year, and this trend is expected to continue. Cloud providers are ramping up construction of AI data centers, such as purchasing large quantities of NVIDIA GPUs, developing in-house chips (e.g., Google TPU, AWS Trainium), and exploring new cooling technologies like liquid cooling.
• Multi-cloud and sovereign cloud go hand in hand: Enterprises tend to adopt multi-cloud deployment to avoid lock-in, but as data regulation tightens, demand for sovereign cloud is rising. Cloud providers are offering localized deployment, data residency, and compliance solutions to meet regulatory requirements across regions for global enterprises.
• Edge computing combines with AI inference: AI inference requires low latency, making edge computing an important part of AI infrastructure. More and more workloads will migrate from centralized data centers to edge nodes, placing new demands on cloud providers' network architecture and edge layouts.
• Sustainability becomes a key indicator: AI data centers consume enormous amounts of energy, making green data centers and renewable energy procurement a key competitive focus for cloud providers. When choosing a cloud provider, enterprises will increasingly consider its ESG (environmental, social, and governance) performance.

These trends indicate that AI is not a short-lived trend in the cloud market, but a catalyst for long-term structural change. In the next five years, AI infrastructure will become the core of global cloud computing industry growth, and the competitiveness of cloud providers will depend on their AI technology depth, infrastructure scale, and ecosystem maturity.

CloudTechDaily Insight

The Q3 cloud market data reveals the start of a new era. AWS's declining market share is, on the surface, a fine adjustment of the competitive landscape, but in essence it is the inevitable result of the cloud market shifting from general-purpose computing to AI computing. Microsoft, leveraging its exclusive partnership with OpenAI, has taken the lead in the AI cloud race; Google, relying on its self-developed TPUs and strong AI R&D capabilities, has held its ground. AWS's posture as a chaser reminds all cloud providers: the moat in the AI era is no longer scale, but technological leadership and the depth of customer lock-in.For enterprise IT strategy, this data means that the choice of cloud service provider will directly determine the success or failure of an enterprise's AI transformation. CISOs and CTOs should evaluate the AI service maturity, data governance capabilities, and long-term roadmaps of each cloud platform, while also paying attention to multi-cloud strategies to avoid excessive lock-in. Although the cost of AI infrastructure is high, it is a necessary investment for enterprises to maintain competitiveness.

Over the next three years, competition in the cloud market will become even more intense, with AI-native cloud, sovereign cloud, and edge AI becoming the main incremental markets. We predict that by 2027, AI-related cloud spending will account for more than 60% of the total growth in the cloud market. Whether AWS's market share can stop declining and recover depends on whether it can catch up in AI infrastructure. This cloud market transformation triggered by AI is just beginning.

CloudTechDaily will continue to track this trend and provide in-depth analysis and insights for enterprise decision-makers.