1. AWS Data Centres Around the World
Amazon Web Services (AWS) continues to expand its global infrastructure as of August 2025. As of 2024, AWS operated 33 global regions comprising more than 100 availability zones across over 20 countries, totaling in excess of 125 physical data centres, with more under construction. [1]
Recent expansions include the establishment of new facilities in Chile, New Zealand, Saudi Arabia and Taiwan, reflecting growing global demand for high-performance computing and AI. [2]
Region-specific investments include:
- Australia: AWS is investing nearly A$20 billion (about US$13 billion) by 2029 to expand data centres in Sydney and Melbourne, along with three solar farms. [3]
- India: AWS plans to invest US$12.7 billion by 2030 to develop cloud and AI infrastructure, reinforcing the country’s tech ecosystem. [4]
- United States: AWS has committed at least US$20 billion in Pennsylvania, and US$10 billion in North Carolina, for AI and cloud infrastructure. [5]
On a company-wide scale, AWS and Amazon are executing a record capital expansion. In 2025, AWS’s infrastructure spending is part of Amazon’s projected US$100 billion annual CapEx, with AI-powered data centres consuming a sizeable share. [5]
Collectively, these developments underline AWS’s ongoing commitment to delivering localised, high-performance services while scaling up to meet AI-driven demand worldwide.
2. Drivers Behind the Expansion
Several interlinked factors are propelling AWS’s global data centre build-out:
Soaring Demand for AI and High-Performance Compute
The explosion of artificial intelligence workloads requires immense computational power. AWS is boosting its supply of Nvidia H100 AI GPUs and building bespoke AI data centres, such as Project Rainier utilising AWS’s own Trainium 2 chips. [2][6]
Strategic Global Presence and Latency Reduction
Expanding into emerging markets such as Latin America (Mexico, Chile), Asia (Taiwan, New Zealand), and the Middle East (Saudi Arabia) helps AWS bring services closer to users, improving speed, reliability and data sovereignty. [2]
Sustainability and Energy Efficiency
AWS is integrating renewable energy solutions and efficiency upgrades into its data centre strategy. For example, solar farms in Australia support the power demands of new facilities. [3] AWS also incorporates innovations in power, cooling and hardware design for more energy-efficient data centres. [7]
Long-Term Vision and Ownership of Technology
AWS is investing in its own chip capabilities, especially with Trainium 2, to reduce cost per performance and decrease reliance on third-party GPU suppliers. [8][4] CEO Andy Jassy emphasises that, although capital outlay comes upfront, data centres deliver returns over many years, making for attractive long-term free cash flow and return on invested capital (ROIC).
Domestic Growth Potential
Despite its current dominance, many businesses are yet to fully transition to the cloud. AWS leadership sees substantial future growth among customers still migrating their operations. [9]
3. Expectations on Return on Investment (ROI)
Long-Term Capital Deployment
Investments in data centres are sizable and up-front. However, AWS treats them as long-lived assets with benefits accruing over 15 to 20 years. The lag between capital spend and revenue generation means returns accumulate over time, delivering strong long-term free cash flow. [8]
Pricing Efficiency Through Hardware Innovation
Trainium 2 chips are said to offer 30–40% better price-performance compared to existing GPU solutions. Combined with improvements in AI inference models, caching and infrastructure, this is expected to lower overall AI costs for customers, which could drive greater adoption and revenue growth. [8]
Client ROI and Commercial Viability
AWS believes that cloud migration, especially with AI included, provides quantifiable returns for clients through agility and reduced infrastructure burden. Customers increasingly insist on ROI-positive AI initiatives rather than novelty, pushing AWS to support measurable outcomes. [9]
Macroeconomic Scale and Cost Reductions
CEO Andy Jassy forecasts that progress in chip efficiency and cloud infrastructure will make AI compute cheaper, unlocking mass adoption. This increased volume of usage is expected to boost AWS revenue faster than increases in cost. [8]
4. Overview Table
| Focus Area | Key Details |
|---|---|
| Global Expansion | New regions in Chile, New Zealand, Saudi Arabia, Taiwan; big investments in Australia, India, US |
| Main Drivers | AI demand, latency, sustainability, chip strategy, growth potential |
| ROI Outlook | Heavy upfront cost, but long-term returns through efficiency and volume growth |
| Customer Impact | Lower AI costs, better ROI, accelerated cloud adoption |
The company’s approach is not short-sighted. Instead, AWS is investing at scale, accepting upfront costs in exchange for durable infrastructure capable of delivering sustained revenue and free cash flow over many years. Innovations like Trainium 2, paired with smarter AI inference and infrastructure design, are helping to reduce the effective cost per AI compute. This strategy is meant to deliver value both to AWS shareholders and to its customers, many of whom are now insisting on tangible ROI from cloud and AI deployment.
Overall, AWS’s data centre and AI infrastructure investments reflect a long-range commitment to powering global digital transformation, where early capital—is turned into widespread adoption, cost-effective services, and reliable returns on investment.
References
- AWS global regions and availability zones. (Dgtl Infra ⧉, Wikipedia ⧉)
- New data centres in Chile, New Zealand, Saudi Arabia, Taiwan. (Top AI Tools List - OpenTools ⧉, AsiaConnect Magazine ⧉)
- Australia investment including solar farms. (The Wall Street Journal ⧉, The Australian ⧉)
- India investment plans. (The Times of India ⧉)
- US investments (Pennsylvania, North Carolina) plus global CapEx scale. (The Wall Street Journal ⧉, DataCenterDynamics ⧉, MarketScreener ⧉)
- Rising demand and power needs for data centres. (Tech in Asia ⧉)
- Sustainable infrastructure initiatives. (Amazon Web Services, Inc. ⧉)
- Chip cost-performance and long-term ROI (Andy Jassy). (Business Insider ⧉)
- AWS CEO on customer ROI and tech adoption. (The Verge ⧉)
