
Nvidia’s $89B data center quarter and AWS’ plan for two million more GPUs show AI infrastructure demand broadening beyond hyperscalers.
Nvidia's data center chip revenue has grown dramatically, and major cloud providers are ordering millions of additional GPUs to meet customer demand for AI workloads. Data center operators face a challenge in building the physical infrastructure to support this compute demand, as the number of GPUs ordered does not directly translate into electrical load requirements. The issue is that physical construction of facilities, power generation, cooling systems, and networking infrastructure cannot keep pace with the speed at which AI technology deployment is advancing. Power availability is already becoming a constraint in many markets, and distinguishing between announced compute capacity and actual executable electrical load is critical for utilities and infrastructure planners.

AI data centers' massive capital investments are at risk from corrosion, which begins during construction – not operations – yet preservation is rarely included in project governance or early planning.

At IFA 2026, Nvidia and its partners showed off the first RTX Spark-powered laptops and mini PCs, designed to run AI models right on your computer.

Power and water aren’t the only constraints on new builds. In many markets, fiber access and diversity now determine where facilities can be located and how quickly they can scale.
Want to go deeper than the news? Explore live, cohort-based AI courses taught by practitioners.
Browse AI courses on Maven