Skip to content

Map

Independent AI Infrastructure: Gigawatt Scaling and the Hyperscaler Leasing Shift

Independent AI data center operators and Bitcoin miners (Core Scientific, TeraWulf, Hut 8, Applied Digital, Cipher Digital, Lambda) signed gigawatt-scale infrastructure leases and credit facilities in April and May 2026.

Updated May 18, 10:00 UTC3 narratives7 confirmed
7 sources • 7 primary

Published by Compute Statecraft. Read the method before treating an inferred claim as confirmed.

Evidence weight

High

7 primary source(s), 0 secondary source(s), 0 social source(s).

Last factual audit

May 18, 10:00 UTC

No explicit correction note is currently visible in the changelog.

Analytical ladder

Level 1

Confirmed fact

Directly supported by listed sources in the Confirmed section.

Level 2

Inference

Reasoned synthesis from multiple facts, made explicit in narrative provenance.

Level 3

Strategic implication

What operators, firms, or regulators may do if the pattern holds.

Level 4

Scenario / watchpoint

Signals that could change the assessment but are not yet proven outcomes.

Confirmed

  • Independent AI data center operators and Bitcoin miners (Core Scientific, TeraWulf, Hut 8, Applied Digital, Cipher Digital, Lambda) signed gigawatt-scale infrastructure leases and credit facilities in April and May 2026.

    SOURCED2 citation(s)
  • Lambda closed a $1 billion senior secured credit facility to expand its AI cloud and data center footprint.

    SOURCED1 citation(s)
  • Core Scientific raised $3.3 billion in notes to fund AI colocation infrastructure and expanded its gross power pipeline to 4.5 gigawatts, including a 1.5 gigawatt plan for its Muskogee campus.

    SOURCED2 citation(s)
  • Hut 8 commercialized the first phase of its 1 gigawatt Beacon Point campus with a 15-year, 352 megawatt colocation agreement designed for NVIDIA infrastructure.

    SOURCED1 citation(s)
  • Applied Digital signed a 15-year lease with a U.S. investment-grade hyperscaler for 300 megawatts of IT load at its Delta Forge 1 campus.

    SOURCED1 citation(s)
  • Cipher Digital signed its third AI data center campus lease with an investment-grade hyperscaler and closed a $200 million revolving credit facility.

    SOURCED1 citation(s)
  • TeraWulf secured a $250 million revolving credit facility and generated $21 million in Q1 HPC lease revenue from 60 megawatts of energized IT capacity.

    SOURCED1 citation(s)

Narratives

Hyperscaler outsourcing

rising

Hyperscalers are signing 15-year colocation leases with independent operators to secure immediate gigawatt-scale power rather than building all infrastructure themselves.

  • These leases concentrate independent-operator revenue in just a few hyperscaler tenants.
  • A hyperscaler shift toward self-build could strand gigawatts of speculative independent capacity.

Supports: 2 • Contradicts: 0 • Context: 1

Power Lens

Capital

  • The capital model for AI infrastructure is shifting from equity to massive debt facilities, as seen in Lambda's $1 billion credit line and Core Scientific's $3.3 billion note offering.

Compute

  • Independent operators have proven they can secure and scale multi-gigawatt power envelopes, making them critical intermediaries for hyperscalers facing grid bottlenecks.

What would change this

  • A sharp reduction in hyperscaler AI capex would disproportionately hit independent operators relying on 15-year leases to service billions in new debt.

  • Regulatory intervention or grid operator limits on gigawatt-scale campus interconnections could stall the announced power pipelines.

Changelog

  • May 18, 10:00 UTC

    Initial publication based on the Q2 2026 financing and leasing wave across Core Scientific, Lambda, Hut 8, Applied Digital, Cipher, and TeraWulf.

Claim ledger

Level 1 - Confirmed factSOURCEDEvidence High

Independent AI data center operators and Bitcoin miners (Core Scientific, TeraWulf, Hut 8, Applied Digital, Cipher Digital, Lambda) signed gigawatt-scale infrastructure leases and credit facilities in April and May 2026.

Level 1 - Confirmed factSOURCEDEvidence Medium

Lambda closed a $1 billion senior secured credit facility to expand its AI cloud and data center footprint.

Level 1 - Confirmed factSOURCEDEvidence High

Core Scientific raised $3.3 billion in notes to fund AI colocation infrastructure and expanded its gross power pipeline to 4.5 gigawatts, including a 1.5 gigawatt plan for its Muskogee campus.

Level 1 - Confirmed factSOURCEDEvidence Medium

Hut 8 commercialized the first phase of its 1 gigawatt Beacon Point campus with a 15-year, 352 megawatt colocation agreement designed for NVIDIA infrastructure.

Level 1 - Confirmed factSOURCEDEvidence Medium

Applied Digital signed a 15-year lease with a U.S. investment-grade hyperscaler for 300 megawatts of IT load at its Delta Forge 1 campus.

Level 1 - Confirmed factSOURCEDEvidence Medium

Cipher Digital signed its third AI data center campus lease with an investment-grade hyperscaler and closed a $200 million revolving credit facility.

Level 1 - Confirmed factSOURCEDEvidence Medium

TeraWulf secured a $250 million revolving credit facility and generated $21 million in Q1 HPC lease revenue from 60 megawatts of energized IT capacity.

Level 3 - Strategic implicationINFERREDEvidence Low

The capital model for AI infrastructure is shifting from equity to massive debt facilities, as seen in Lambda's $1 billion credit line and Core Scientific's $3.3 billion note offering.

Level 3 - Strategic implicationINFERREDEvidence Low

Independent operators have proven they can secure and scale multi-gigawatt power envelopes, making them critical intermediaries for hyperscalers facing grid bottlenecks.

Level 4 - Scenario / watchpointSPECULATIVEEvidence Low

A sharp reduction in hyperscaler AI capex would disproportionately hit independent operators relying on 15-year leases to service billions in new debt.

Level 4 - Scenario / watchpointSPECULATIVEEvidence Low

Regulatory intervention or grid operator limits on gigawatt-scale campus interconnections could stall the announced power pipelines.

Source Library

primary sources

secondary sources

No sources listed.

social sources

No sources listed.