Why Executives Are Rejecting Cloud Dependency and Turning to AI Factories

According to Intersect360 Research data, sovereign AI data centers have become the fastest-growing segment of accelerated computing infrastructure as C-suite executives reject public cloud dependency and pivot toward on-premises, high-performance integrated architectures. Spending is projected to reach $122.8 billion by 2026.

According to data from Intersect360 Research and the HPC-AI Leadership Organization, sovereign AI deployments stand out as high-performance computing centers operating under high-density artificial intelligence. While 62.5% of organizations participating in the research run traditional high-performance computing and AI workloads on the same systems, 67.1% use the same shared storage tier for both workloads.

In sovereign AI initiatives such as Japan's AI Bridging Cloud Infrastructure project or the G42 cloud in the United Arab Emirates, keeping critical workloads away from the public cloud is enforced as a mandatory rule. Research data indicates that 56.3% of sites keep their active computing infrastructure entirely on-premises, while 68.2% follow the same approach for storage.