F'back: Call with Delta Electronic's Director of DC Solutions & Architecture

研报英文原文证据摘录

F'back: Call with Delta Electronic's Director of DC Solutions & Architecture

EUROPE | Clean Technology EquityMayResearch20, 2026

F'back: Call with Delta Electronic's Director of

DC Solutions & Architecture

AI data centre power constraints are shifting the industry toward

new architectures and supply-driven decisions. Grid interconnection

delays, transient load challenges and severe equipment bottlenecks

are accelerating adoption of 800V DC systems, SSTs and behind-the-

meter generation. Near-term technology choices are primarily dictated by

availability and speed to deployment, rather than cost, with supply chain

constraints the dominant gating factor.

We hosted Jason Agee, Director of DC Solutions and Architecture at Delta Electronics.

Power constraints are driven by grid limitations and transient load dynamics - The primary

constraint for AI data centres has shifted from simple power availability to the grid’s ability

to handle highly volatile, transient demand. AI training workloads create sudden inrush loads,

requiring extensive grid stability studies and controlled ramp-up periods to avoid system

failures. This has made interconnection timelines unpredictable and prolonged, particularly

as even approved projects can still cause instability in practice. The result is that “where and

when can I get power” is now the central question for developers. Consequently, time-to-power

has overtaken cost as the dominant decision metric, fundamentally reshaping investment

prioritisation.

Supply chain bottlenecks dominate deployment timelines across the stack - Supply chain

constraints are pervasive across all layers of data centre power infrastructure, from generation

through to distribution. Long lead times are visible across transformers, SSTs, fuel cells and