Cost & Opex · Efficiency Metrics

PUE / WUE

Power Usage Effectiveness and Water Usage Effectiveness — global and regional baselines, company-reported figures, cooling-technology tradeoffs, and regulatory targets.

Confidence A Directly reported figure B Estimated / Derived C Regional inference / indirect estimate D Not found / insufficient data

Industry-Wide Average PUE · Global 1.52

Power Usage Effectiveness, lower is better · range across tracked baselines shown below

<1.3 · Best-in-class, top 23% (lowest/best)~2.1 · EU enterprise data centers (highest)

PUE baselines tracked 6 global, EU, regulatory target

Directly-reported (Conf. A) 4 / 6 remainder estimated/derived

PUE Trend

PUE — global & regional baselines

BASELINES

Scope	PUE	Conf.
Industry-wide average (global)	1.52	A
Largest facilities (global)	1.45	A
Best-in-class (top 23% of respondents)	<1.3	A
EU average (JRC estimate)	~1.6	A
EU enterprise data centers	~2.1	B
Regulatory target (2030)	<1.4	B

PUE — company-reported

COMPANY DATA

Company	Type	PUE	Conf.
Google	Hyperscaler	~1.09	B
Meta	Hyperscaler	~1.09	B
AWS	Hyperscaler	~1.15	B
Equinix	Colocation	~1.39	B
Digital Realty	Colocation	~1.58	B

WUE — by scope

BY SCOPE

Scope	WUE (L/kWh)	Conf.
Industry average (global)	1.8–1.9	B
Ideal (air-cooled, no evaporative)	0.0	A
Microsoft Arizona facilities	1.52	A
Microsoft Singapore facilities	0.02	A
Amazon (company-wide average)	0.19	A
EU CNDCP commitment (water-stressed sites)	≤0.4	A

Cooling technology — WUE vs. PUE tradeoff

TRADEOFF

Cooling technology	WUE (L/kWh)	PUE	Conf.
Liquid (direct-to-chip)	Near 0	1.05–1.2	B
Evaporative	1.5–3.0	15–30% improvement vs. mechanical-only	B
Best-in-class (NREL reference facility)	0.7	1.06	B