The framework in one paragraph

The AI Power Chain runs from the utility service territory boundary to the point-of-load conversion at the GPU or accelerator die. The six intermediate domains cover grid interconnection, facility distribution, rack-shelf rectification and wide-bandgap conversion, on-package power delivery, thermal management, and modular build. Vendor concentration, supply-chain lead times, standards trajectory and qualification cycles all vary across them; none is predictable from the others. Sponsors, operators and hyperscaler strategy teams that treat the six as one commodity input systematically under-price recovery risk in specific domains and over-invest in others.

Why the framework exists

AI infrastructure discussion tends to collapse around three visible actors: chip companies, hyperscalers, and data centre operators. The physical infrastructure between the grid meter and the accelerator package runs invisibly under those three, treated as commodity input.

That treatment held when rack power sat below 30 kW and facility distribution standardised on 415V AC. It has stopped holding as rack power crossed 100 kW during the 2024-2025 GB200 NVL72 deployments, and as facility distribution began to fragment between legacy 415V AC, the OCP sidecar architecture (Mt. Diablo v0.7.0, with a default ±400V bipolar output and a permitted 800V unipolar option), and the native MVAC-to-800V VDC architecture on the Nvidia Blackwell and Rubin reference platforms and OCP's longer-term facility path. At that point the invisible infrastructure becomes a set of six distinct engineering problems, each with its own vendor set, its own qualification cycle, and its own lead-time exposure. The vendor consolidation at thermal management (Eaton acquiring Boyd Thermal in March 2026 for approximately $9.5B per company disclosures, plus Ecolab acquired CoolIT from KKR for approximately $4.75B (closed 2 July 2026) per Ecolab disclosures) is being priced by strategic buyers who already read the market this way. Independent sponsors and asset-level equity holders are largely still pricing the domains as commodity.

The framework applies the same analytical tradition as classical value-chain decomposition in industrial-tech investing, updated for the specific supply-chain, thermal and standards constraints of accelerator-dense compute above 100 kW per rack. It was first developed for internal use during commercial + technical DD workstreams in 2024-2025 and codified in the architecture map essay in early 2026.

The six domains

Framework calibration

Case studies apply the physical scorecard alongside the Investment Layer verdict on 2024-2026 transactions. Each transaction stresses different domains of the framework, which is the practical test of whether the six-domain decomposition survives contact with real deals.

• AirTrunk / Blackstone + CPP (Sep 2024). Per-domain scores put physical at ~22% (LOW-MEDIUM band). The clean-asset calibration anchor; low physical concentration across all six domains.
• Crusoe / Stargate Abilene. Physical HIGH on grid interconnection (behind-the-meter gas) and rack shelf (single-accelerator SiC concentration). The framework's grid-and-generation scenario, materialised.
• CoreWeave DDTL 3.0. Portfolio-averaged physical MEDIUM. Compute delivered from colocation footprint rather than greenfield; the debt-facility calibration path.

What the framework tells you

The framework produces three analytical claims that individual essays develop in depth. The current Vendor Screen indicates vendor concentration is highest at grid interconnection, rack shelf (wide-bandgap) and thermal management, and lowest at modular build. Value capture concentrates unevenly across the six domains, which the operator-level narrative that treats them as commodity fails to price.

Lead-time exposure is highest at grid interconnection (200+ week transformer lead times) and thermal management (multi-year qualification cycles for coolant chemistry and quick-disconnects). Rack shelf and on-package delivery have supply-side capacity growing on a two-to-four year investment cycle. Facility distribution lead-time exposure is currently short but concentrating around the 800V DC transition. Modular build lead-time exposure is driven by skilled-labour availability rather than equipment supply.

Standards fragmentation is highest at facility distribution, where 415V AC, sidecar 800V DC and native 800V DC architectures coexist, at on-package delivery (proprietary per-accelerator interfaces), and at modular build (limited multi-vendor interface standards). Sponsors and operators making architectural decisions in fragmented domains commit to specific vendor stacks for the useful life of the equipment, and the cost of choosing wrong compounds over the full asset life.

The AI Power Chain Vendor Screen scores 300+ vendors across the six domains on moat, chokepoint, AI exposure and execution. The scoring is the operational output of the framework applied to the vendor universe.

What the framework does not do

The framework decomposes the physical infrastructure into six engineering domains. It surfaces where value concentrates and where lead-time and standards risks compound. Asset-level, domain-specific financial modelling requires separate machinery and belongs in the Investment Layer essays.

The framework is a snapshot of the physical infrastructure architecture as of August 2026. It covers physical infrastructure only. Software above the accelerator, model-training economics that drive accelerator demand, and customer-side capacity procurement decisions belong to adjacent frameworks that were developed and applied separately.

The framework treats the six as sequential in the physical power path but not necessarily sequential in the operator or sponsor analytical process. Real DD workstreams cover the six in parallel with heavy interaction between facility distribution and thermal management, and between grid interconnection and modular build. The numbering follows physical scope. Analytical priority is separate.

Where the framework has been applied

The framework structures the analytical content across the AI Power Chain series, the Investment Layer series, the Due Diligence for the AI Buildout series, and the DC-DC Transition series. The three cross-cutting essays inside the AI Power Chain series explicitly work across all six layers:

• The AI Power Chain Vendor Screen, 300+ vendors scored across all six domains
• Pricing Under Scarcity, where premium capture is compounding across the domains
• The Services Inversion, where services multiples are converging with product multiples across the domains

The Investment Layer series applies the framework to capital allocation decisions:

• The AI infrastructure M&A map, 2023-2026 transaction pattern mapped to the six domains
• The PE playbook, layer by layer, sponsor economics by domain
• Underwriting AI infrastructure debt, recovery risk by domain

The Due Diligence for the AI Buildout series operationalises the framework across sixteen DD workstreams, with several workstreams (product architecture and technology DD, manufacturing, IP and supply chain DD) explicitly organised by domain.

Framework evolution

Three open questions currently shape how the framework will evolve through 2027-2028. Each will be revisited as the underlying market data resolves.

The facility-distribution architecture question resolves as the OCP-standardised interfaces are adopted across both the sidecar (Mt. Diablo, with configurable bipolar or unipolar output) and the native MVAC-to-800V VDC path. The durable split is retrofit vs greenfield rather than vendor vs vendor. Rack shelf and on-package delivery vendor concentration follow the facility-distribution outcome with a two-to-three year lag.

The thermal-versus-modular boundary is dissolving as thermal capacity is increasingly integrated into modular prefab skids at the vendor level (Boyd/Eaton, Motivair/Schneider). Whether the two remain analytically separable at the framework level is an open question and will be revisited if the operational integration compounds.

The grid-versus-facility boundary is also under pressure as behind-the-meter power procurement (nuclear PPAs, on-site generation, hyperscaler-utility JVs) transitions from one-off transaction into repeatable strategy. The framework may absorb some of the grid-interconnection scope into a new "generation" domain if the pattern compounds.