Michigan Campus AI Data Center Expansion targets an initial 52 megawatts (MW) of AI-focused computing capacity as owner Hyperscale Data advances plans to transform its Michigan facility into a large-scale artificial intelligence infrastructure hub.
The project is expected to begin with a 20 MW deployment of critical power capacity, with the first 10 MW targeted to become operational within 90 days and a second 10 MW phase expected approximately three months later. An additional 32 MW of capacity is planned for delivery in 2028, bringing the campus’s initial buildout to roughly 52 MW.
To support the expansion, Hyperscale Data plans to transition the site away from Bitcoin mining operations and dedicate available power resources to AI computing, GPU deployments, and colocation services. The company said growing demand for scalable AI infrastructure has increased interest in the Michigan campus from prospective customers seeking high-density compute environments.
The Michigan campus currently supports digital infrastructure operations and serves as the foundation for the planned AI expansion. According to the company, the site has the long-term potential to support more than 300 MW of total power capacity through future development phases, subject to regulatory approvals, financing, utility agreements, infrastructure availability, and engineering studies.
If realized, the 52 MW buildout would represent only a portion of the campus’s potential capacity, leaving substantial room for future expansion as demand for AI computing infrastructure continues to grow.
Part of the broader U.S. AI infrastructure buildout
The Michigan AI Data Center Campus Expansion is part of a broader wave of AI infrastructure development taking shape across the United States. Another notable project is the Nakota AI Data Campus, which recently secured a strategic partnership with Nidar Infrastructure Limited and Yotta Data Services, including an anchor customer commitment for its first 100 MW phase. The project targets an initial 100 MW deployment. And could support up to 1 GW of capacity upon full buildout as developers pursue large-scale AI computing and data center investments.
Michigan Campus AI Data Center Expansion: Project factsheet
Project Name: Michigan Campus
Location: Michigan, United States
Owner: Hyperscale Data, Inc.
Operator: Alliance Cloud Services (ACS), a wholly owned subsidiary of Hyperscale Data
Project Type: AI Data Center / Colocation Infrastructure
Planned Initial Capacity: 52 MW
Phase 1 Capacity: 20 MW
- First 10 MW targeted within 90 days
- Additional 10 MW targeted approximately 90 days later
Future Expansion Capacity: 32 MW planned for 2028
Total Initial Buildout: 52 MW
Long-Term Capacity Potential: More than 300 MW (subject to approvals, financing, utility agreements, and infrastructure availability)
Current Site Use: Digital infrastructure operations and Bitcoin mining
Planned Site Transition: Phased conversion from Bitcoin mining to AI computing and colocation services
Primary Workloads:
- Artificial Intelligence (AI)
- High-density GPU computing
- AI training and inference
- Enterprise IT infrastructure
- Colocation services
Revenue Potential (Company Estimate):
- 20 MW deployment: More than $1 billion over 20 years
- 52 MW deployment: Approximately $2.5 billion over 20 years
Project Status: Development / Customer onboarding phase
Key Milestone:
- Initial 20 MW AI deployment targeted to begin following execution of a customer services agreement.