Technical Manual · City Profiles
DC Markets — City Deep-Dive Methodology
Metric definitions, data provenance, and profile structure behind the 10-city global data center market hub. Each profile is a data-driven editorial reference — metrics are drawn from regulator publications, CBRE/JLL market reports, and Uptime Institute surveys. No computed quantities are fabricated; ranges reported in the source literature are preserved as ranges here.
▶ Open the DC Markets hub01 Purpose & basis
The DC Markets hub is a curated reference for data center professionals evaluating global market entry, site selection, or competitive benchmarking. It covers ten major markets: Singapore, Frankfurt, London, Dubai, Jakarta, Kuala Lumpur, Mumbai, Northern Virginia, Sydney, and a hub index page that compares all ten side-by-side.
Each city profile is a static editorial snapshot — a structured data card, not a live data feed or interactive calculator. The page presents published market intelligence (capacity, efficiency, power cost, connectivity, regulatory posture, tenant mix) in a consistent schema so professionals can compare markets on common dimensions. No user inputs are accepted; no equations are executed client-side. The value of the hub is curation, structure, and source citation — not computation.
Scope note: profiles describe colocation and wholesale data center markets serving enterprise and hyperscale tenants. Edge, telco-exchange, and single-tenant enterprise campuses are excluded unless they materially define a market's capacity base (e.g., hyperscaler campuses in Northern Virginia).
02 Profile fields — what each page shows
No user computation occurs. The "inputs" to each profile are the city selection (which page you open) and the static data fields compiled for that market. The table below documents every displayed field, its unit, and the data-basis category.
| Field | Unit | Display section | Data-basis category |
|---|---|---|---|
| Operational capacity | MW | Key Metrics | Operator disclosures + CBRE/JLL market reports |
| Average PUE | dimensionless ratio | Key Metrics | Market survey average — Uptime Institute Global DC Survey + regulator guidance |
| Power cost | USD/kWh | Key Metrics | National tariff schedule or IEA country data; commercial/industrial rate band |
| Colocation price | USD/kW/month | Key Metrics | CBRE/JLL asking-rate ranges for standard colocation rack power; median of published range |
| Vacancy rate | % | Key Metrics | CBRE/JLL availability ratio at time of report |
| Annual growth rate | % | Key Metrics | Year-on-year installed-capacity growth, CBRE/JLL or operator pipeline disclosures |
| Applicable standards | list | Key Standards | Certification frameworks active in the market; verified against regulator and operator disclosures |
| Major operators | list | Major Operators | Named colocation or wholesale providers with disclosed presence; sourced from operator websites and market reports |
| Cooling strategy | prose | Cooling Strategy | Editorial synthesis of dominant approaches per climate zone and operator practice |
| Key challenges | prose list | Key Challenges | Editorial; anchored to regulatory filings, operator earnings commentary, and analyst reports |
| FAQ answers | prose | FAQ accordion | Market-specific questions; answers cite the regulatory or market source inline |
| References | numbered list | References section | Primary sources: regulator URLs, CBRE/JLL report pages, IEA data |
03 Metric definitions & data basis
Precise definitions for each headline KPI, with the data-basis tag and source provenance used to populate the profiles.
Operational capacity (MW)
Operational capacity = sum of commissioned IT-load capacity across
colocation + wholesale facilities in the market
Measured in megawatts of available IT load (not gross building power draw). Does not include announced-but-unbuilt pipeline. Sources: operator disclosures (earnings reports, press releases), aggregated by CBRE Data Centre Solutions or JLL Data Centre Research. Where published figures vary, the more conservative (lower) estimate is used.CBRE/JLL · MARKET
Power Usage Effectiveness (PUE)
PUE = Total facility power / IT equipment power
= (IT load + cooling + power distribution + lighting + other) / IT load
Definition per ISO/IEC 30134-2 and Green Grid White Paper #49. Values shown are the market average of reported annual PUE across active facilities — not a single-facility figure. Market averages are drawn from the Uptime Institute Global Data Center Survey (published annually) supplemented by national regulator benchmarks where available (e.g., Singapore IMDA requires new DC PUE ≤ 1.3; Germany EU CoC baseline; EU Energy Efficiency Directive Art. 12 disclosures). A lower PUE indicates higher energy efficiency; 1.0 is the theoretical minimum.Uptime Institute · ISO/IEC 30134-2 · STANDARD
Power cost ($/kWh)
Power cost = commercial/industrial electricity tariff at the grid connection point
Expressed in USD/kWh at the medium-voltage (MV) or high-voltage (HV) tariff band applicable to large commercial facilities (typically >1 MVA demand). Does not include renewable energy premium, wheeling charges, or PPA uplift unless stated. Currency conversion uses approximate annual-average exchange rates. Sources: national energy regulators (EMA Singapore, Bundesnetzagentur Germany, Ofgem UK, DEWA Dubai, PLN/RUPTL Indonesia, TNB Malaysia, MSEDCL/Tata India, Dominion Energy / PJM USA, AER Australia).National regulators · IEA · MARKET
Colocation price ($/kW/month)
Colo price = contracted rack-power rate, standard colocation (per-kW billing)
Reflects the asking rate or market median for standard colocation (not hyperscale wholesale) at the time of the report. Actual contracted rates depend on committed power draw, contract term, redundancy tier, and operator. Published as a single midpoint where CBRE/JLL report a range (e.g., $350–$430/kW/mo → $390). Sources: CBRE Asia Pacific Data Centre Trends, JLL Data Centre Outlook, Cushman & Wakefield Data Centre Market Reports.CBRE/JLL/CW · MARKET
Vacancy rate (%)
Vacancy rate = unleased operational capacity / total operational capacity × 100
Availability ratio for powered colocation or wholesale space that is commissioned, fit-out-ready, and available for lease. A low vacancy rate (e.g., <2%) indicates a supply-constrained market. Sources: CBRE/JLL market reports (quarterly or bi-annual).CBRE/JLL · MARKET
Annual growth rate (%)
Annual growth rate = (current year capacity − prior year capacity) / prior year capacity × 100
Year-on-year change in commissioned IT-load capacity. Where pipeline data is used in lieu of actual commissioned figures (fast-growing markets), the figure is an estimated absorption rate based on CBRE/JLL pipeline schedules. Treat as a directional indicator, not a precise forecast.CBRE/JLL · EDITORIAL
Fiber connectivity count
Fiber count = number of distinct submarine cable systems landing or terminating
in the market (at any cable landing station)
Sourced from TeleGeography SubmarineCablemap.com and operator IR materials. Terrestrial dark-fiber diversity (carrier-neutral meet-me rooms, carrier hotels) is described qualitatively in the Cooling & Connectivity section. Higher fiber count correlates with lower cross-border latency and carrier redundancy.TeleGeography · MARKET
Regulatory overlay
Regulatory overlay = set of binding rules specifically governing DC operations
(beyond general commercial building codes)
Examples: Singapore IMDA DC Green Mark moratorium + PUE mandate; Indonesia RUPTL power allocation; Germany IT-Sicherheitsgesetz; UAE TDRA Tier requirements; Australia DCCEEW Guarantee of Origin. Each profile's Standards section lists the applicable certification frameworks; the FAQ accordion addresses the dominant regulatory question for that market.National regulators · EDITORIAL
04 Per-city data provenance
The primary source used for each city's headline metrics. Where multiple sources were consulted, the most authoritative or most recent publication is listed first. All references appear in full in each city profile's References section.
| City | Capacity & growth source | Power cost source | PUE basis | Regulatory anchor |
|---|---|---|---|---|
| Singapore | CBRE APAC; JLL Singapore DC Outlook | EMA Singapore tariff schedule | Uptime Institute Survey; IMDA PUE mandate ≤ 1.3 (new) | IMDA DC Green Mark; BCA Green Mark; SS 564 |
| Frankfurt | CBRE EMEA; JLL EMEA DC Outlook | Bundesnetzagentur; Eurostat industrial tariff | Uptime Institute Survey; EU Code of Conduct baseline | EU Energy Efficiency Directive Art. 12; BSI IT-Grundschutz |
| London | CBRE EMEA; Cushman & Wakefield UK DC | Ofgem; UK BEIS electricity price statistics | Uptime Institute Survey; EU CoC (pre-Brexit adoption retained) | CDEI; Planning Policy; ISO 27001 prevalent |
| Dubai | CBRE MENA; JLL Dubai DC | DEWA tariff (commercial >100 kVA) | Uptime Institute Survey; TDRA Tier standards | TDRA; eSafety regulations; UAE PDPL |
| Jakarta | CBRE APAC; JLL Indonesia DC; APJII | PLN tariff (B-3 / I-4 band); RUPTL | Uptime Institute Survey; limited public disclosure | RUPTL power allocation; GR 71/2019 (data localization) |
| Kuala Lumpur | CBRE APAC; JLL Malaysia DC | TNB tariff (C2/C3 commercial high-voltage) | Uptime Institute Survey; MCMC DC guidelines | MCMC; PDPA 2010; MyDigital Blueprint DC targets |
| Mumbai | CBRE India; JLL India DC Outlook | MSEDCL/Tata Power HT tariff | Uptime Institute Survey; BEE energy efficiency norms | BEE; IT Act 2000 (amended); India DPDP 2023 |
| Northern Virginia | CBRE Americas; JLL N. Virginia DC; Dominion Energy reports | Dominion Energy commercial tariff; PJM capacity charges | Uptime Institute Survey; EPA ENERGY STAR DC benchmarks | Virginia data center tax incentives; NERC reliability standards |
| Sydney | CBRE APAC; JLL Australia DC | AER; AEMO wholesale; Ausgrid/Endeavour commercial tariff | Uptime Institute Survey; DCCEEW DC energy benchmarking | DCCEEW Guarantee of Origin; Privacy Act 1988; ASD ISM |
| Hub index | Aggregated from above; CBRE/JLL global comparative | IEA World Energy Prices (cross-country) | Uptime Institute Global DC Survey (market medians) | Editorial summary; links out to per-city profiles |
Data vintage: figures in the profiles reflect publications available at time of authorship (primarily 2024–2025 report cycles). Market conditions — especially power cost, colocation pricing, and vacancy rates — can change within a single quarter. Treat all figures as illustrative benchmarks rather than live transactional data.
05 Profile card outputs
What a reader gets from each city profile page, and how to interpret each field.
| Output field | How displayed | Interpretation guidance |
|---|---|---|
| Operational capacity (MW) | Hero metric card | Commissioned IT load in the market. Compare across cities to gauge market scale. Does not include announced pipeline. |
| Average PUE | Hero metric card | Market mean; individual facilities may differ ±0.2 or more. A PUE of 1.0 = perfect efficiency; 1.5 = 50% overhead. Lower is better. |
| Power cost ($/kWh) | Hero metric card | Commercial grid tariff rate. Combined with PUE, drives total energy cost per MW of IT load. |
| Colo price ($/kW/mo) | Hero metric card | Published asking rate for standard colocation. Hyperscale wholesale rates are typically 30–50% lower and are not shown here. |
| Vacancy rate (%) | Hero metric card | Markets below 2% vacancy are supply-constrained; expect longer lead times and upward price pressure. |
| Annual growth rate (%) | Hero metric card | Year-on-year capacity expansion. Fast-growing markets may have supply catching up — check vacancy alongside growth. |
| Applicable standards | Icon list | Certification frameworks active in the market. Uptime Institute Tier and TIA-942 are globally portable; local standards (SS 564, BCA) are market-specific contractual requirements. |
| Major operators | Card grid | Named providers with disclosed operational presence. Not an exhaustive directory; hyperscaler captive campuses are generally excluded. |
| Cooling strategy | Prose callout | Dominant approach for the climate zone. Dictates the PUE floor achievable without premium technology (e.g., liquid cooling). |
| Key challenges | Numbered prose list | Structural barriers to capacity expansion or operations. Ordered roughly by impact severity. |
| FAQ accordion | Expandable Q&A | Three market-specific questions most commonly asked by professionals evaluating the market. Answers cite primary sources. |
| References | Numbered citation list | Primary sources for the profile data. Follow these for due diligence or deeper research. |
06 Worked example — Singapore profile
Walk through the Singapore city profile as a representative example: how each field is derived, what the figures mean operationally, and how the regulatory context shapes the numbers.
- Operational capacity — 850 MW: CBRE APAC and JLL Singapore DC Outlook (2024 cycle) aggregate commissioned colocation and wholesale capacity across Jurong, Tuas, and central carrier-hotel locations. Pipeline (approved under the 2022 post-moratorium allocation) adds approximately 350 MW of announced-but-uncommissioned capacity, bringing the headline "1.2 GW" figure cited in the hero description. The metric card shows the operational 850 MW; the 1.2 GW refers to total addressable including near-term pipeline. Operational = 850 MW
- Average PUE — 1.55: Singapore's tropical equatorial climate (28–32 °C year-round, 70–90% RH) forces chillers to run continuously at high ambient wet-bulb temperatures. The Uptime Institute 2024 Global Survey places the Asia-Pacific median PUE at approximately 1.55, consistent with Singapore operator disclosures. New facilities under the 2022 Green Mark scheme must target PUE ≤ 1.3, driving liquid cooling and indirect evaporative pre-cooling adoption in new builds. Market average PUE = 1.55
- Power cost — $0.18/kWh: EMA Singapore publishes quarterly non-contestable tariffs; the commercial high-tension band (HTB) applicable to data center grid connections was approximately SGD 0.23–0.25/kWh in 2024, translating to roughly USD 0.17–0.19/kWh at prevailing exchange rates. The profile uses $0.18 as the midpoint. Power cost = $0.18/kWh
- Colocation price — $390/kW/mo: CBRE APAC Data Centre Trends (H1 2024) quotes Singapore standard colocation asking rates at $350–$430/kW/month for a typical 10–20 kW rack cabinet commitment. The profile reports the arithmetic midpoint ($390). Colo price = $390/kW/mo
- Vacancy — 0.8%: With moratorium constraints on new supply and strong hyperscaler demand, Singapore has maintained sub-1% vacancy since 2022. CBRE APAC reports 0.8% as of H2 2023. This signals a severely supply-constrained market — expect 12–24-month lead times for new deployments. Vacancy = 0.8%
- Annual growth — 12%: Post-moratorium approvals (2022–2024) have accelerated pipeline deliveries. JLL Singapore DC Outlook (2024) estimates 12% annualised capacity growth over the 2023–2025 window, driven by approved hyperscaler and colo expansions. Annual growth = 12%
- Moratorium context (FAQ): The 2019–2022 moratorium halted new builds while Singapore's DCs consumed ~7% of national electricity. The 2022 partial lift introduced strict conditions: PUE ≤ 1.3, Green Mark Platinum DC certification, mandatory waste-heat reporting, and capacity caps per allocation round. This regulatory structure explains the gap between the 850 MW operational and 1.2 GW pipeline figure — the pipeline is allocated but subject to phased commissioning gated on efficiency evidence. Regulatory context documented in profile FAQ
Reading note: the six KPI cards on each profile page are comparable across cities on the same dimension. However, do not derive secondary quantities by combining metrics across different profile pages without verifying that the underlying definitions align — for example, colocation pricing in Northern Virginia reflects a very different contract structure and power cost environment than Singapore, so a simple ratio comparison overstates or understates the true cost delta.
07 References & standards bodies
- Uptime Institute — Global Data Center Survey (annual): market-average PUE, operational practices, tier distribution. Primary PUE benchmark for all 10 profiles.
- CBRE Data Centre Solutions — Asia Pacific Data Centre Trends; EMEA Data Centre Market Report; Americas Data Centre Market Report: capacity, vacancy, colocation pricing, growth rates.
- JLL Research — Data Centre Outlook (regional): pipeline, hyperscaler activity, operator moves, rental ranges. Used as secondary corroboration to CBRE figures.
- IEA World Energy Prices — Cross-country electricity price comparison (commercial/industrial tariff band). Used for power cost cross-check where national regulator data was ambiguous.
- TeleGeography SubmarineCablemap.com — Submarine cable system counts and landing station locations per market. Primary source for fiber connectivity figures.
- Singapore: IMDA + EMA + NEA + BCA — Green DC Roadmap (PUE mandate), EMA tariff schedule, climate impact, Green Mark scheme. City-level sources for Singapore profile.
- Indonesia: PLN RUPTL + APJII — Power allocation framework, tariff bands (B-3/I-4), ISP infrastructure data. City-level sources for Jakarta profile.
- Malaysia: TNB + MCMC — Commercial HV tariff (C2/C3), DC guidelines, MyDigital Blueprint. City-level sources for Kuala Lumpur profile.
- UAE: DEWA + TDRA — Commercial tariff (>100 kVA), Tier requirements, eSafety regulations. City-level sources for Dubai profile.
- India: MSEDCL / Tata Power + BEE — HT industrial tariff, energy efficiency norms. City-level sources for Mumbai profile.
- Germany: Bundesnetzagentur + EU CoC — Industrial electricity tariff, EU Code of Conduct for Data Centres baseline PUE. City-level sources for Frankfurt profile.
- UK: Ofgem + BEIS — Electricity price statistics, grid carbon intensity. City-level sources for London profile.
- USA: Dominion Energy + PJM + EPA — Commercial tariff, capacity market charges, ENERGY STAR DC benchmarks. City-level sources for Northern Virginia profile.
- Australia: AER + AEMO + DCCEEW — Wholesale and retail tariff, Guarantee of Origin scheme, DC energy benchmarking. City-level sources for Sydney profile.
- ISO/IEC 30134-2 — PUE definition standard. The Green Grid White Paper #49 provides the engineering elaboration used in this manual.
08 Assumptions & limitations
The DC Markets hub is an independent personal research resource derived from publicly available sources. Key limitations to carry when using the data:
Data vintage. Capacity, pricing, and vacancy figures reflect published market data from approximately the 2023–2025 report cycle. Data center markets in Southeast Asia, the Middle East, and India are growing at 15–35% per year; figures can be materially different within six months. For procurement or site-selection decisions, verify against the most recent CBRE/JLL quarterly update.
Market averages vs. individual facilities. PUE is a market average; individual facilities range from 1.2 (best-in-class liquid cooling in temperate climate) to 2.0+ (legacy air-cooled tropical). Colocation pricing is the median asking rate; hyperscale wholesale transactions can price 30–50% lower with long-term power commitments.
Currency conversion. All prices are stated in USD. Local-currency tariffs are converted at approximate annual-average exchange rates and may diverge from spot rates by ±10–15%.
Operator presence. The Major Operators list is representative, not exhaustive. Operators not listed may have presence in a market; listed operators' footprints change through acquisitions and expansions. Verify with the operator's own market page for current availability.
Regulatory timeliness. Regulatory status (moratoriums, green mandates, data-localization requirements) can change with little public notice. The profiles document the regulatory situation at time of authorship; always confirm with the relevant authority having jurisdiction before committing to a deployment.
Disclaimer: ResistanceZero is independent personal research and does not represent any current or former employer. Content is for educational and informational purposes only and is not investment, legal, or procurement advice. See Terms & Disclaimer.