According to ABI Research Principal Analyst Sebastian Wilke, the U.S remains at the forefront of AI data center investments
The demand for AI-focused data centers is set to grow significantly in the coming years, outpacing general-purpose cloud infrastructure by a factor of two to three, according to Sebastian Wilke, principal analyst at global technology intelligence firm ABI Research.
“We see demand for AI-focused data centers to grow significantly over the next few years, driven by several converging trends,” Wilke said. “We expect rapid expansion, intense competition among GPU cloud providers, as well as increasing investor and provider focus on efficiency, location, and differentiation, the ABI Research analyst added.
One of the key drivers for this expected expansion is the explosive growth of AI workloads. “Large language models, computer vision and real-time inference require massive compute power,” the analyst explained. “This is pushing hyperscalers and AI-native startups alike to build or partner with specialized AI infrastructure providers.”
ABI Research also noted that this growing complexity is also fueling a shift toward more specialized hardware. “GPUs, and custom AI silicon (like GCP’s TPU, AWS Trainium/Inferentia or Azure’s Maia) are replacing general-purpose CPUs in many workloads,” Wilke noted. “In turn requiring data centers optimized for high-density compute, advanced cooling, and power efficiency.”
Beyond performance and hardware, geopolitics and regulatory frameworks are reshaping the AI data center landscape. “Countries and enterprises are seeking control over where data is processed, stored and models are trained, mainly because of regulatory fragmentation and geopolitical strategic rivalry,” the ABI Research analyst said. “This is prompting regional investments in localized, AI-optimized data centers.”
In parallel, demand is spreading beyond hyperscale hubs. “In addition to hyperscale growth, edge and micro data centers are being developed to support real-time AI at the network edge,” Wilke said. “Expect a surge in workload-optimized micro-clouds serving finance, pharma, defense, and media. The next AI wave won’t be centralized—it will be strategically distributed.”
Sustainability concerns are another mounting pressure for AI infrastructure developers. “As energy consumption grows exponentially for AI data centers and we are at a point we see diminishing returns from hyper-effective cooling technologies, these dedicated data centers will face increasing scrutiny to adopt environmental practices,” Wilke warned. “This is leading to innovation in liquid cooling, renewable integration, location decision and AI-driven energy optimization.”
Regionally, the U.S remains at the forefront of AI data center investments. “The U.S remains the epicenter of AI data center investment, with coastal clusters like Northern Virginia and Silicon Valley dominating model training, while regions such as Ohio, Texas, and Utah are booming due to favorable power, tax and land economics,” the ABI Research Analyst said.
Meanwhile, Europe is balancing its digital ambitions with regulation. “Europe is focusing on AI centers tied to regulatory compliance, EU digital autonomy, and data residency, even at the cost of efficiency,” Wilke said. “The EU is thereby desperately attempting to balance both regulation and incentives for local innovation.”
“The Middle East, particularly the UAE and Saudi Arabia, is emerging as somewhat of a wildcard investor, with almost unlimited capital, low energy costs, and ambitions to be AI compute hubs of the Global South.”
The ABI Research analyst also noted that Asia-Pacific is poised for the fastest growth. “Across APAC (excluding China), rapid growth is expected through 2030, with the region projected to host the largest number of colocation sites globally,” Wilke said. “Countries like Malaysia and Thailand are attracting significant investment from hyperscalers, driven by advancements in cloud orchestration, low latency requirements and rising regional AI demand.”
“China continues to pursue a parallel AI infrastructure path, driven by massive state investment and self-reliance goals as a result of U.S. export controls of the past years,” he added. “Despite external constraints, internal demand and support from hyperscaler like Alibaba, Tencent and Huawei sustain its momentum.”
Wilke also noted broader investment trends. “Across all regions, rising power demands and carbon concerns are prompting investment in renewable-powered and liquid-cooled data centers,” he said. “While in the U.S. AI data center growth is largely driven by private tech giants, China, the Middle East, and Europe lean more heavily on public-private partnerships and state-led initiatives.”
