Base case · strong conversion
Base: strong conversion
Constrained: power / capex bind
How hyperscaler AI capex converts into Eaton and Trane power-and-cooling order backlogs
A vertical waterfall from AI model demand and hyperscaler capex through a data-center hall with a power lane and a cooling lane, gated by grid availability into Eaton and Trane backlogs, then pulled into revenue by book-to-bill. A scenario toggle throttles the conversion.
1
2
3
4
5
6
7
AI model training
inference demand
Hyperscaler capex
surge
Data center construction
power + cooling per MW
Eaton — power distribution
Trane — applied cooling
MV switchgear
transformer
PDU
chiller
chilled-water loop
AHU / CRAH
GPU racks
power + cooling meet at the chip
Grid / power availability
gates live MW
power binds · capacity delayed
Eaton electrical
backlog
Trane applied
commercial-HVAC backlog
Book-to-bill above 1.0
pulls segment revenue through
Swing factor (2026-2028)
capex normalization / power constraints
strong conversion