800 VDC Data Center Power Architecture Market to Reach USD 59.76 Billion by 2034 as AI Racks Outgrow AC Power

The global 800 VDC data center power architecture market was valued at $16.24 billion in 2025 and is projected to reach $59.76 billion by 2034, a 15.6% CAGR, as AI rack power demand drives adoption. Brownfield retrofits accounted for 58.36% of 2025 deployments, with sidecar systems expected to support staged upgrades and native 800 VDC revenue expected around 2029. Safety standards, DC arc-flash risks and retrofit costs remain challenges; Asia Pacific is projected to grow fastest, at a 16.7% CAGR.

Analysis

The investable signal is not the long-dated market forecast; it is which suppliers can monetize the transition phase by phase. Brownfield sidecars favor vendors able to integrate protection, conversion, backup power and controls into existing halls. That may support Eaton (ETN), ABB (ABBN) and Vertiv (VRT), but the article provides no order, backlog or revenue evidence to distinguish them. Eaton–Infineon (IFX) creates a specific qualification pathway for silicon-carbide content, while broad semiconductor exposure at IFX or Navitas (NVTS) should not be confused with material earnings sensitivity. Over time, native DC could reduce some legacy conversion equipment and copper per delivered MW, even as higher AI capacity offsets those reductions; winners depend on total power buildout, not architecture share alone.

Near term, treat this as a low-impact thematic catalyst: a market-research forecast and standards activity do not establish customer adoption. Over 1–3 months, watch named vendor orders, customer qualification and safety certification. Over 6–18 months, the key uncertainty is whether deployments advance beyond pilots and sidecars toward repeatable projects. Grid interconnection constraints can defer equipment demand regardless of efficiency gains. DC arc-flash rules, competing voltage standards, retrofit cost, or delayed rack launches could reverse the adoption narrative. The contrarian point: the market may overstate near-term supplier revenue by treating a 2034 forecast as a current order pipeline; conversely, verified retrofit wins could be underappreciated because native DC remains later-dated.

Key Decisions for Investors

• Do not chase a broad 800 VDC basket on this release alone. The forecast is not independently verified demand, and the article supplies no vendor-specific order or earnings data.
• Watch for a conditional relative-value long in ETN versus a broad industrial benchmark if management reports repeatable data-center power orders tied to retrofit projects and maintains guidance. Reassess if orders/backlog weaken or certification and deployment timelines slip.
• Track IFX as a qualification-driven semiconductor exposure, not a clean 800 VDC pure play. Require evidence that the Eaton platform moves from partnership to production volumes before assigning meaningful earnings upside; monitor relevant segment disclosures and customer qualifications.
• Use vendor disclosures over the next 1–3 months as the catalyst check: confirmed deployments, backlog conversion, and safety approvals support the thesis; pilot-only activity, delayed rack launches, or persistent grid-connection delays falsify the near-term demand case.