Methodology.
Directories disagree about what counts as a data center by a factor of three. This is where we draw the line, where our data comes from, and how we keep it deduplicated across sources.
§2 · What we map
A facility makes the map only if it passes all five tests.
§3 · Inclusion matrix
| Category | Status | Notes |
|---|---|---|
Commercial colocation | Include | Equinix, Digital Realty, CoreSite, DataBank… |
Hyperscale buildings | Include | Each named building, not just the campus |
Enterprise (≥ 500 kW) | Include | Banks, retail, healthcare — when documented |
Carrier hotels with tenants | Include | One Wilshire, 60 Hudson, Telehouse… |
Standalone HPC centers | Include | NCSA, ORNL, etc. Purpose-built and at scale |
Standalone edge facilities | Include | Vapor IO, Compass Edge — actual buildings |
Cloud regions | Include | Tracked as a separate layer on the map |
Government (when public) | Include | DoE/NASA-published, etc. |
Pure telco central offices | Exclude | Switching only — not a data center |
Crypto mining facilities | Exclude | Different category, different operating model |
University CS server rooms | Exclude | Research compute is not infrastructure-as-product |
Cabinets at cell towers | Exclude | Edge compute ≠ a facility |
Server rooms in non-DC buildings | Exclude | Closet test fails |
Rumored or unannounced sites | Exclude | Wait for a source |
§4 · Where the data comes from
PeeringDB
The authoritative directory for interconnect-relevant facilities. Operators voluntarily list themselves to make peering arrangements easier.
OpenStreetMap
Crowd-tagged buildings with telecom=data_center. 210 observed, 115 deduped against PeeringDB. Useful for facilities outside the interconnect ecosystem.
Operator websites
Equinix, Digital Realty, DataBank, Cologix, CoreSite, CyrusOne, and QTS publish per-facility spec pages. These add power capacity, cabinet density, UPS topology, certifications, and other operational details that PeeringDB doesn't capture.
Cloud provider region pages
AWS, GCP, Azure, Oracle. Tracked as a separate map layer, not in the facility table. A region is a logical grouping of buildings, not a building.
§5 · How we match and dedupe
Each canonical facility row in the database can have many source records linked to it — one for every directory or page where we found that facility. When a new source mentions a facility, we try to attach it to an existing canonical row before creating a new one. The match function checks (operator, name) for an exact match first, then any facility within 100m. If neither hits, the record is logged as an orphan and reviewed before insertion.
This is why an Equinix facility known by three different names across PeeringDB, OSM, and Equinix’s own website still shows up as one row, with three source records linked to it.
§6 · Corrections and submissions
Find an error? Know about a facility we’re missing? Send the details — building name, operator, address, and a verifiable source — and we’ll add it.
Operators: if you’d like your facility list verified directly rather than scraped, get in touch.
Data from PeeringDB (CC-BY-SA), OpenStreetMap (ODbL), and operator-published facility pages · Map tiles by Mapbox