AI computing power rises from North America to Europe

Since the beginning of 2026, the global cloud computing market has undergone a historic shift. Major international cloud providers, including Amazon Web Services (AWS), Google Cloud, and Microsoft Azure, have successively raised prices on AI-compute-related products, breaking the traditional logic of the cloud computing industry—which had long been driven by economies of scale leading to price reductions. From North America to Europe, and from chips to storage, the entire AI computing power supply chain is experiencing a systemic price increase cycle.

Behind this round of price hikes lies the combined effect of surging demand for computing power, driven by the explosion in AI applications, and rigid constraints on the supply side. As the dual-engine of "training + inference" becomes the norm and AI agents like OpenClaw spark deployment frenzies, the global supply-demand imbalance for computing power has become irreversible.

Panorama of Price Hikes: Widespread Increases from North America to Europe

1 North American Cloud Giants Lead the Price Hikes
In January 2026, Amazon Web Services (AWS) broke with industry convention by announcing a roughly 15% price increase for its EC2 Machine Learning Capacity Blocks.
Google Cloud followed suit. In March, Google Cloud announced price adjustments for its data transfer services, including CDN Interconnect, Direct Peering, and Carrier Peering, effective May 1st. These adjustments included increases of up to 100% in North America, 42% in Asia, and 60% in Europe.
While Microsoft Azure did not directly announce an across-the-board price increase, the costs of its AI services have been passed on to end-users through various channels. According to market research data, the actual transaction prices for Azure AI services in Q1 2026 rose approximately 8%–12% compared to the previous quarter.

2 European Market Follows Suit
Outside North America, the European cloud services market is also experiencing price surges. Major European cloud providers, such as Germany's Hetzner and France's OVHcloud, have successively announced price increases, typically ranging from 5% to 15%. The rationale behind the European increases mirrors that of North America—soaring demand for AI computing power, coupled with tight hardware supply chains, has forced service providers to adjust their pricing strategies.

It is worth noting that while the magnitude of price increases in Europe is less pronounced than in North America, the transmission has been faster. Due to Europe's lack of large-scale domestic AI chip manufacturing capabilities, its reliance on imported hardware is higher, making it more directly susceptible to supply chain cost fluctuations.

3 Price Transmission Across the Entire Industry Chain
Price increases are not limited to cloud services; they are propagating upstream along the industry chain. From chip design and manufacturing to packaging and testing, and from memory components to printed circuit boards (PCBs), the entire AI hardware supply chain has entered a price upswing.

In the memory chip sector, prices for core components like DRAM, High Bandwidth Memory (HBM), and Solid State Drives (SSDs) continue to climb. Chey Tae-won, Chairman of SK Group, stated at NVIDIA's GTC conference in March 2026 that surging AI demand is driving a comprehensive recovery in demand for DRAM, HBM, CPUs, and SSDs, with orders now requiring reservations months or even years in advance.
The packaging and testing industry is also experiencing upward price momentum. Capacity for advanced packaging-related services, such as wafer-level packaging and bumping, has tightened. Coupled with rising raw material costs, pricing power in the packaging and testing sector is strong in 2026. In passive components, the usage of Multi-layer Ceramic Capacitors (MLCCs) in AI servers can reach up to eight times that of traditional servers, and structural price increases have already begun.