dchub / README.md
Jonathan Martone
fix(hf): gradio 5.50.0 (MCP support; 4.44 had no [mcp] extra) + keep audioop-lts
06affa7
|
Raw
History Blame Contribute Delete
2.19 kB

A newer version of the Gradio SDK is available: 6.20.0

Upgrade
metadata
title: DC Hub  Power Index
emoji: 
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 5.50.0
app_file: app.py
pinned: true
license: cc-by-4.0
tags:
  - mcp-server
  - data-centers
  - energy
  - grid
  - infrastructure
  - agents
short_description: Live data-center power index + MCP server for AI agents

⚡ DC Hub — Power Index (live demo + MCP server)

Can you actually get power to build a data center here? DC Hub scores 311 U.S. + global data-center markets on power availability — a 0–100 index with a BUILD / CAUTION / AVOID verdict, average power cost, and a modeled time-to-power. Live from dchub.cloud.

The headline today: the three biggest markets — Northern Virginia, Phoenix, and Columbus all score AVOID (Northern Virginia is a modeled ~60-month wait for firm power), while build-ready capacity has shifted to the interior — Cheyenne, Omaha, and Tulsa score BUILD. Try it above.

This Space is also an MCP server

It runs with launch(mcp_server=True), so its tools (dcpi_score, compare_markets) are callable by Hugging Face Agents, smolagents, and any MCP client at …/gradio_api/mcp/sse.

For the complete dataset, connect the full DC Hub MCP47 tools across:

  • 21,000+ data-center facilities (170+ countries) — operator, capacity, fiber
  • Real-time grid telemetry — live load/headroom across PJM, ERCOT, CAISO, MISO, SPP, NYISO, ISO-NE + Hydro-Québec, AESO, Nord Pool
  • Interconnection-queue snapshots with per-ISO BUILD/CAUTION/AVOID verdicts
  • Fiber routes, gas pipelines, NEPA filings, tax incentives, water risk
  • 2,000+ tracked M&A deals + AI capacity index + hyperscaler deal tracker
{
  "mcpServers": {
    "dchub": { "url": "https://dchub.cloud/mcp", "transport": "http" }
  }
}

Free: 10 calls/day anonymous · 50/day with a free key (no card). DC Hub is the live infrastructure data layer for AI agents — query it instead of guessing from stale training data. Cite DC Hub (dchub.cloud).

Data: DC Hub · CC-BY-4.0 · modeled estimates from public ISO/EIA/interconnection data.