Speed-to-Power for
AI Inference at Scale

Inference compute online in months, not years. Nodiac deploys modular data centers at existing power sites with distribution-level interconnects, bypassing the 5+ year IX queue. 500+ sites identified across 30+ states — the portfolio below shows our top-scored sites.

1–15 MW per siteDistribution-level interconnectHyperscaler-readyPowered-shell or turnkey
Total Data Center Capacity
available in months
Total Data Center Sites
across 0 states
Base Average Data Center
portfolio-wide
Project Areas of Interest
Regions and Local Zones

Portfolio Map

Where the power is.

Loading map...

Nodiac Sites

● Hover sites for more information
● Click map to zoom, click outside to scroll page
Map Legend
Nodiac Site Locations
125+ GW Opportunity Sites
Size = Estimated Capacity
Site MW
1 MW
5 MW
10 MW
20 MW
50 MW
100+ MW
Major Sites
50 MW
100 MW
150 MW
200 MW

State Scorecard

Loading live data...

Weighted Capacity Index (Conservative) vs. State

Hover over bars for detailed state information

Loading chart data...
Top State (WCI)
Loading...
Top by High MW
Loading...
Top Projects State
Loading...
Loading scorecard data...

WCI: A confidence-weighted score based on the MW we can reliably bring online, adjusted for hyperscaler presence and connectivity.

How We Score: Proximity signals: Regions, AWS Local Zones, Azure/GCP Edge POPs, NOVA adjacency. Status weights: Yes = 1, Planned = 1, Likely = 0.5. Weights: Region ×5; AWS Local Zone ×1; Azure/GCP Edge POP ×1; NOVA adjacency ×3 per provider (MD/DC).

Scoring Formulas: Proximity = (Σ status-weighted Regions × 5) + (LZ × 1) + (Edge POPs × 1) + (NOVA × 3 per provider). WCI = Conservative MW × (1 + 0.15 × Proximity + 0.05 × Fiber) + Conservative Average Project.

Fiber: 0 = Sparse/rural, 1 = Strong secondary metros, 2 = Core hubs (dense interconnection).

Why Nodiac

Traditional data center builds are power and timeline constrained. We place modular infrastructure where the power already is.

1

Speed to Capacity

Distribution-level interconnects shorten critical paths for 1–15 MW increments.

2

Resilient & Lower-Carbon

Co-located with renewables and near dispatchable resources for 24×7 supply.

3

De-risked Execution

500+ sites screened. Proven playbook from $26B+ in energy project development.

4

We Handle the Hard Part

Land, interconnection, power, permitting, and operations. You bring the GPUs.

How It Works

1

Site Selection

Screen existing power sites for capacity, interconnect availability, and hyperscaler proximity.

2

Power Connection

Secure partnership with site owner. Handle permitting, utility agreements, and interconnection.

3

Deploy Infrastructure

Ship and install factory-built data center modules. Connect power, cooling, and network.

4

Inference Compute Online

Commission, validate, and bring infrastructure to hyperscaler-ready spec. Scale as demand grows.

Commercial Models

1

Powered Shell

Nodiac delivers interconnect, switchgear, power & cooling. Partner brings racks and network.

2

Turnkey Colocation

White space, power, cooling, and operations included. Consume GPU hours or lease capacity.

3

JV Structure

Strategic deployments with shared economics and long-term roadmap alignment.

Methodology & Assumptions

Portfolio Inputs

• State-level aggregation from Fleet Sites Validated data.
• Project size range directly from Nodiac site information data.
• Averages computed as High MW ÷ project count.
• Provider presence sourced from a curated Provider Catalog.

How We Score

Proximity signals: Regions, AWS Local Zones, Azure/GCP Edge POPs, NOVA adjacency.
Status weights: Yes/Planned = 1.0, Likely = 0.5, No = 0.0
Fiber tiers: 2 = Core hubs, 1 = Secondary metros, 0 = Sparse/rural.

Scoring Formulas

Proximity = (Σ status-weighted Regions × 5) + (LZ × 1) + (Edge POPs × 1) + (NOVA × 3 per provider)
WCI = Conservative MW × (1 + 0.15 × Proximity + 0.05 × Fiber) + Conservative Average Project

Request a Technical Briefing

We'll walk through interconnect paths, grid imports, site screening, and commercial options tailored to your workloads.