Dataset Coverage
This page shows how complete the public dataset is across 344 tracked facilities. It doubles as a research backlog: every missing field here is a gap to close in the index. For raw exports, visit Data Downloads.
Coverage Snapshot
Core Fields
| Field | Why it matters | Present | Missing | Coverage |
|---|---|---|---|---|
| Operators | Operator and ownership grouping | 344 | 0 | 100% |
| Country | National taxonomy and search intent | 344 | 0 | 100% |
| Region | Subnational geography | 344 | 0 | 100% |
| City | City-level lookup and local reporting | 344 | 0 | 100% |
| Status | Buildout timeline and feed readiness | 344 | 0 | 100% |
| Capacity | Scale, rankings, and aggregate totals | 344 | 0 | 100% |
| Coordinates | Map, globe, and geographic analysis | 344 | 0 | 100% |
| Sources | Verifiability and citation quality | 344 | 0 | 100% |
| Energy Type | Power sourcing and carbon analysis | 344 | 0 | 100% |
| AI Focus | Training, inference, sovereign, and cloud segmentation | 344 | 0 | 100% |
Timeline Readiness
The site can only support year, feed, and milestone pages if entries carry consistent event fields. Right now, `start_year` is the most established timeline field; milestone-specific years are still sparse.
What this unlocks next
- Year archive pages such as `/year/2026/` and `/year/2025/`
- Status transition feeds for operators and countries
- Milestone trend charts for announced, construction, and live capacity
- Faster QA on stale entries that never moved beyond announcement
Advanced Reporting Fields
| Field | Present | Coverage |
|---|---|---|
| `investment_usd` | 344 | 100% |
| `cooling_type` | 344 | 100% |
| `water_context` | 344 | 100% |
| `grid_impact` | 344 | 100% |
| `hardware` | 344 | 100% |
These are the highest-value underreported dimensions in the directive: investment, cooling, water, grid stress, and hardware stack. As they fill in, they can graduate into their own public taxonomy pages.
Current Gaps to Prioritize
Missing Capacity
0 entries still lack a disclosed MW or GW number, which limits rankings and global totals.
Missing Coordinates
0 entries still lack map-ready latitude/longitude, which weakens the globe and geographic views.
Missing Energy / AI Focus
0 entries lack `energy_type` and 0 lack `ai_focus`, weakening taxonomy pages added in recent iterations.
Coverage by Country
Best-Documented Countries
Countries with at least 3 tracked facilities and the strongest field completeness across capacity, coordinates, energy, AI focus, and timeline.
| Country | Facilities | Readiness |
|---|---|---|
|
Denmark
Cap 100% · Geo 100% · Time 100%
|
3 | 100% |
|
China
Cap 86% · Geo 100% · Time 100%
|
7 | 97% |
|
South Africa
Cap 86% · Geo 100% · Time 100%
|
7 | 97% |
|
United Kingdom
Cap 100% · Geo 100% · Time 89%
|
9 | 96% |
|
Nigeria
Cap 75% · Geo 100% · Time 100%
|
4 | 95% |
|
Egypt
Cap 67% · Geo 100% · Time 100%
|
3 | 93% |
|
South Korea
Cap 100% · Geo 100% · Time 82%
|
11 | 89% |
|
Ireland
Cap 100% · Geo 100% · Time 80%
|
5 | 88% |
|
France
Cap 100% · Geo 100% · Time 78%
|
9 | 87% |
|
United States
Cap 100% · Geo 100% · Time 69%
|
84 | 86% |
Largest Country Research Gaps
These countries already have enough tracked facilities to matter, but their metadata is still thin enough to constrain rankings, feeds, or deeper trend pages.
| Country | Facilities | Readiness |
|---|---|---|
|
Kuwait
Cap 100% · Geo 100% · Time 0%
|
3 | 40% |
|
Taiwan
Cap 100% · Geo 100% · Time 20%
|
5 | 52% |
|
New Zealand
Cap 100% · Geo 100% · Time 25%
|
4 | 55% |
|
Norway
Cap 100% · Geo 100% · Time 20%
|
5 | 60% |
|
Philippines
Cap 100% · Geo 100% · Time 33%
|
3 | 60% |
|
Qatar
Cap 100% · Geo 100% · Time 33%
|
3 | 60% |
|
Russia
Cap 100% · Geo 100% · Time 33%
|
3 | 60% |
|
Australia
Cap 100% · Geo 100% · Time 50%
|
8 | 70% |
|
Mexico
Cap 100% · Geo 100% · Time 50%
|
4 | 70% |
|
Turkey
Cap 100% · Geo 100% · Time 50%
|
4 | 70% |
Coverage by Operator
Best-Documented Operators
Operators with at least 3 tracked facilities and the strongest documentation density across the core fields that drive taxonomy and trend views.
| Operator | Facilities | Readiness |
|---|---|---|
|
EuroHPC JU
Cap 100% · Energy 100% · Time 100%
|
6 | 100% |
|
Alibaba Cloud
Cap 100% · Energy 100% · Time 100%
|
4 | 100% |
|
Eviden (Atos)
Cap 100% · Energy 100% · Time 100%
|
4 | 100% |
|
Vantage Data Centers
Cap 100% · Energy 100% · Time 100%
|
4 | 100% |
|
Amazon
Cap 100% · Energy 100% · Time 100%
|
3 | 100% |
|
Apple
Cap 100% · Energy 100% · Time 100%
|
3 | 100% |
|
AWS
Cap 100% · Energy 100% · Time 100%
|
3 | 100% |
|
ByteDance
Cap 100% · Energy 100% · Time 100%
|
3 | 100% |
|
Naver
Cap 100% · Energy 100% · Time 100%
|
3 | 100% |
|
Oracle
Cap 100% · Energy 100% · Time 91%
|
11 | 98% |
Largest Operator Research Gaps
Useful targets for the next content passes: operators with enough footprint to rank and compare, but not enough field completeness to fully expose their buildout.
| Operator | Facilities | Readiness |
|---|---|---|
|
Africa Data Centres
Cap 0% · Energy 100% · Time 100%
|
4 | 80% |
|
Cassava Technologies
Cap 0% · Energy 100% · Time 100%
|
4 | 80% |
|
Digital Realty
Cap 100% · Energy 67% · Time 67%
|
3 | 80% |
|
Microsoft
Cap 100% · Energy 79% · Time 73%
|
56 | 86% |
|
AMD
Cap 100% · Energy 100% · Time 33%
|
3 | 87% |
|
Hewlett Packard Enterprise
Cap 88% · Energy 100% · Time 50%
|
8 | 88% |
|
Amazon Web Services
Cap 100% · Energy 84% · Time 81%
|
43 | 89% |
|
Google
Cap 100% · Energy 86% · Time 83%
|
42 | 90% |
|
NVIDIA
Cap 71% · Energy 100% · Time 90%
|
21 | 92% |
|
Meta
Cap 100% · Energy 89% · Time 89%
|
19 | 94% |
Related Surfaces
Coverage complements global statistics, data downloads, and the operator/country taxonomy pages. For readers comparing facility hardware claims with real local inference benchmarks, SiliconBench is the most relevant sibling property.