QumulusAI (QMLS) Registration filing summary

Company overview and business model

• Specializes in rapid deployment of GPU-powered cloud infrastructure for AI, targeting underserved small and mid-market customers as well as large enterprises needing flexible, high-performance compute solutions.

• Operates over 1,120 GPUs across three U.S. data centers, with expansion plans exceeding 120 MW of IT load and potential to support over 90,000 NVIDIA B200/B300 GPUs.

• Revenue streams include AI/HPC compute services, blockchain mining, and hosting, with a strategic shift toward HPC as the primary growth driver.

• Utilizes a dual-channel go-to-market strategy: marketplace partners (e.g., RunPod) and direct enterprise sales, optimizing for both scale and customer intimacy.

• Recent acquisition of The Cloud Minders, Inc. (TCM) diversifies revenue and enhances recurring, stable income from GPU-based infrastructure.

Financial performance and metrics

• 2025 revenue was $11.85M, up 46% from $8.10M in 2024, driven by compute power and hosting services.

• Net loss improved to $4.46M in 2025 from $13.18M in 2024, reflecting operational scaling and acquisition impacts.

• Adjusted EBITDA for 2025 was a loss of $2.35M, down from a $0.59M loss in 2024, due to increased operating costs for public readiness and personnel.

• As of December 31, 2025: total assets $91.7M, cash $11.7M, total liabilities $26.4M, and accumulated deficit $37.5M.

• Significant customer concentration: two customers accounted for 67% of 2025 revenue.

Use of proceeds and capital allocation

• No proceeds will be received from the direct listing; all shares sold are by existing shareholders.

• Recent $45M senior secured convertible notes facility and a $500M non-recourse credit facility support working capital, infrastructure expansion, and general corporate purposes.

• $16.5M from the sale of a joint venture interest will be used to accelerate HPC business development.