#What steps is Hyperscale Data taking to expand its services?
Hyperscale Data is actively pursuing plans to lease AI computing capacity from its Michigan Data Center Campus. This move marks the company's transition from primarily serving as a crypto mining operation to becoming an infrastructure-as-a-service provider. Hyperscale Data is assessing long-term growth possibilities for a facility that has the potential to utilize up to 340 MW of power.
#How does the Michigan Data Campus support this transformation?
The Michigan campus of Hyperscale Data encompasses 600,000 square feet and extends over 34.5 acres. Currently, the facility operates at approximately 30 MW of capacity, primarily generating revenue from digital asset mining and colocation services. Expansion objectives are set to increase this capacity to 70 MW by the second quarter of 2027, eventually aiming for a complete buildout of 340 MW by the third quarter of 2029. To support this ambitious growth, the company secured agreements in February 2025 that provide access to 300 MW from an electric utility and an additional 40 MW from a natural gas provider.
#What is the role of the GPU cloud offering?
In tandem with these developments, Hyperscale Data is expanding its GPU cloud service. A new on-demand NVIDIA GPU cloud platform is expected to launch in the first half of 2026, which will offer access to NVIDIA H100, B200, and B300 models specifically designed for AI training, inference, and high-performance computing tasks. Notably, Hyperscale Data already manages NVIDIA GPU clusters for a current cloud provider based in Silicon Valley.
#How is Hyperscale Data financing its expansion?
To raise the necessary capital for development, Hyperscale Data has filed a $125 million shelf registration. This strategy is integral to enabling the growth trajectory outlined by the company.
On the partnership side, Hyperscale Data has collaborated with AGIBOT PTE. LTD., focusing on AI and robotics operations. The company also plans to enhance its workforce by hiring over 500 employees within the next three years.
#What should investors consider about these developments?
Investors ought to remain aware of the associated risks. The filing of a $125 million shelf registration suggests potential dilution in the future. Transitioning from 30 MW today to 340 MW by the third quarter of 2029 carries significant execution risks, particularly regarding construction timelines and actual occupancy of the new capacity with paying clients. It is prudent for investors to monitor any announcements regarding anchor tenants, contract terms, and revenue generated from the GPU cloud services after its launch in 2026.