On June 4, Amazon announced plans to invest $10 billion to build a new data center in North Carolina to expand its artificial intelligence (AI) and cloud computing infrastructure.
The investment is intended to anchor new compute-intensive workloads and help expand Amazon’s capabilities to support enterprises building their businesses with AI, according to a company statement.
Amazon is developing AI software for humanoid robots that could eventually take on delivery tasks, The Information reported, citing anonymous sources. Amazon has reportedly built an indoor obstacle course called "Humanoid Park" at one of its San Francisco offices to test these robots.
Amazon also reached a multi-year licensing agreement with The New York Times at the end of May to bring the latter's news reports, recipes and sports content into Alexa and its proprietary AI models for training.
Why do tech giants need data centers?
Amazon is not the only tech giant that needs to build new data centers. OpenAI announced a super data center project called "The Stargate Project" in early 2025, planning to invest $500 billion over the next four years to build a new artificial intelligence infrastructure in the United States to support the development of OpenAI.
OpenAI stated in its external press release: This infrastructure will ensure the United States' global leadership in artificial intelligence, create hundreds of thousands of American jobs, and bring huge economic benefits to the world. The project will not only promote the re-industrialization of the United States, but also provide a strategic capability to protect the national security of the United States and its allies.
According to previous reports by PowerBeats, on June 3, Meta also signed a 20-year nuclear power procurement agreement with Constellation Energy to meet the growing power needs of data centers.
The reason why technology giants need data centers to build artificial intelligence (AI) and cloud computing infrastructure is that they rely on a large amount of computing resources, storage space, and efficient network connections to process, analyze, and store massive amounts of data. Data centers can provide powerful computing power, large-scale data storage and management, efficient data transmission, and low latency.
1. Powerful computing power
AI algorithms, especially deep learning models, require a lot of matrix operations and data processing. These tasks require extremely high computing power, which is difficult to meet with ordinary computers or servers. Data centers are equipped with high-performance servers and GPU clusters that can provide the necessary computing power to support complex AI training and reasoning processes.
2. Large-scale data storage and management
AI systems usually need to access and process huge data sets, including but not limited to unstructured data such as images, videos, and text. Cloud computing provides elastic and scalable storage solutions, allowing enterprises to dynamically adjust storage capacity according to actual needs while ensuring data security and reliability.
3. Efficient data transmission and low latency
For application scenarios with high real-time requirements (such as self-driving cars, online games, etc.), fast response is crucial. Through optimized internal network architecture and geographical location selection, data centers can achieve high-speed data transmission and extremely low latency to ensure that user experience is not affected.
Data centers provide indispensable infrastructure support for AI and cloud computing, enabling them to ensure performance while also having good economy and flexibility.
In addition, the construction of data centers brings economic and employment benefits to the local area, which is also a byproduct of the construction of data centers.
"Amazon's investment is one of the largest in our state's history and will bring hundreds of high-paying jobs and economic growth to Richmond County," said North Carolina Governor Josh Stein.
Amazon is investing $10 billion to build the infrastructure needed to support large-scale AI models as part of its competition with other tech giants like Google, Microsoft and Meta. Amazon said the funds "will support future AI development from AWS's data centers in the Tasmanian state" and create at least 500 high-skilled jobs in the process.
Amazon will fund technical training programs at community colleges, STEM education at K-12 schools, and career paths for fiber-optic broadband infrastructure. In addition, a $150,000 Richmond County Community Fund has been launched to support the implementation of local workforce development, sustainability, and public health projects.
Conclusion
Leo Fan, co-founder of blockchain-based AI infrastructure company Cysic, told Decrypt: “The expansion of AI infrastructure is good news for the industry, but it also highlights a key problem in this industry: cost. The $10 billion investment shows the high cost of building and expanding AI infrastructure, which prevents smaller developers or enterprises from participating due to lack of necessary infrastructure and hardware funds, thus inhibiting innovation.”
Fan argues that while this investment brings financial benefits, it could also lead to “all innovative AI work being concentrated in the hands of big tech companies,” which he argues could hinder broader innovation.
Amazon may provide the computing power resources required for large AI models to small and medium-sized enterprises through a data center procurement model. However, the centralized computing power supply and demand situation may face obstacles in the instant and flexible acquisition of computing power resources required in the AI era. Decentralized cloud computing power driven by AI running on the blockchain may be a more suitable model for the future.
