January 14th, 2026 – Stifel IRIS is pleased to release “Data Centres: Beyond Hyperscalers,” a comprehensive analysis of the AI-driven capital supercycle and the emerging opportunities for specialists within the data infrastructure ecosystem.
The generative AI boom has triggered an unprecedented expansion in data centre capacity, with global capex projected to reach $389 billion by the end of 2025, a 2.9x increase compared to 2023 levels. While the "Big 3" hyperscalers (AWS, Azure, and Google Cloud) consolidate their 67% share of the public cloud, the sheer scale of the "Compute Era" is creating structural gaps at the periphery. As the industry collides with physical limits, most notably the "Power Wall", the focus is shifting from brute-force training to specialised, inference-driven applications where agility and local mastery are the new prerequisites for performance.
Our latest white paper provides a strategic framework for the "Compute Era," mapping the transition from platform dominance to physical infrastructure control. We analyse the rise of AI-native neoclouds that compete on "goodput" (successful computation) rather than raw price-per-hour, and we examine how Europe is pivoting away from the U.S.-led training race toward sovereign, localised inference. By unbundling the digital infrastructure stack, we identify the "hidden champions" weaponizing local market DNA to carve out high-value moats in the shadow of tech behemoths.
To learn more about these industry trends and see our comprehensive mapping of the European AI infrastructure ecosystem, download the full white paper.
 |
Antoine Lebourgeois
|
