A 14+ layer geospatial atlas of AI/large-load infrastructure:

A 14+ layer geospatial atlas of AI/large-load infrastructure: (1) a curated project workbook ('Project Keystone') of AI-infrastructure campuses with normalized public_status across eight stages (Operational → Announced/early planning → Canceled), a Project Completion Probability score (PCP-1.1.0) as a second dimension, and full project dossiers across overview, infrastructure, diligence, evidence, power, environment, site and market with ranked advantages/constraints and linked queue IDs; (2) 11 proprietary screening indices on a 0.25° US land grid (CSS-4.0.0 site score 0–100 plus climate, water, permitting, tariffs, fiber, behind-the-meter, grid-shock and state-attractiveness layers) with cell research dossiers and factor breakdowns; (3) a hex-based electricity demand layer carrying annual MWh, average MW, sector mix and 2030 AI load projections; (4) on-demand parcel research ranking a top-100 shortlist per cell joined to power and fiber signals

Open buyer room →

Sample

Brickroad customers

Sample this dataset before you buy it.

Your sourcing agent asks Aixenergy and files the sample in your catalog — private to you.

Start 7-day trialNothing is charged until the trial ends.

Coverage

Universe, instruments and categories.

Industry: Financial Exchanges & DataElectric Utilities
Instruments: equities
Categories: Alternative

Dataset card

Type: not stated
Format: Multimodal geospatial — point/polygon map layers, a regular 0.25° raster grid of scored cells and hex-binned demand surfaces, plus per-feature dossier records (structured attributes plus evidence citations) delivered through an interactive map with filterable fields (status, company, geography, business model, evidence, PCP band, MW, capex) and shareable URLs; project data is generated from a workbook through an explicit public-field whitelist build script
Volume: 14+ published layers: one curated project atlas (row count not printed on the catalogue page) carrying 8-stage status plus PCP scores and full dossiers; 11 proprietary screening indices on a 0.25° US land grid (order of 10^5 scored land cells nationally) with per-cell dossiers; a hex demand surface with sector mix and 2030 projections; parcel research returning a top-100 shortlist per cell. A separate public layer ingests the PNNL IM3 OSM atlas of existing US facilities
Users: Est. <100 — a 1-person consultancy/publisher operating a Ghost-based site with paid Sign In / Sign Up portals and an interactive map plus a 'Large Load Monitor' product; audience is a narrow set of energy-finance and development professionals
History: 4 years
Update frequency: not stated
Growth: Active — model versions are explicitly incrementing (CSS-4.0.0 site score, PCP-1.1.0 completion probability), the catalogue was built by a named author set with recent 2026 journal posts on the AI power buildout, and separate map and monitor products are live on their own subdomains
Launched: ~2025 — the journal's imagery and posts are dated 2026, the founder's book 'The Cognitive Grid' is 2025 (with 'The Wind Power Story' 2019), and the company record is newly created; the intelligence map is the 2025–2026 product
Delivery: not stated
Entity mapping: not stated
Sample: not stated
Point-in-time: not stated
Licence: not stated

supplier index read from their site