QumulusAI (QMLS) Investor presentation summary

Market opportunity and demand drivers

• AI infrastructure demand is surging, with capacity as the main constraint, not demand itself.

• Big Tech AI capex is projected to exceed $600B in 2026, up from $390B in 2025.

• US datacenter power demand is expected to triple by 2030, reaching over 180 GW.

• Global datacenter build-out is forecasted at $7T through 2030.

• 90% of notable frontier AI models were produced in 2025, and GenAI adoption reached 53% of the population in three years.

Business model and competitive advantages

• Provides rapid, flexible access to NVIDIA GPUs via colocation and sub-50MW data centers, enabling deployment in months.

• Offers enterprise-grade capacity with both reserved and on-demand GPU pools.

• Delivers predictable, competitive pricing and end-to-end infrastructure control with built-in SLAs and privacy.

• Modular infrastructure allows for fast provisioning and expansion, reducing innovation wait times.

• Smaller, distributed sites enable faster market entry, lower delivery risk, and easier permitting.

Growth, traction, and financial performance

• Deployed NVIDIA GPU fleet grew 5.5x since mid-2025, reaching 3,088 units.

• Q1 2026 revenue grew 83% YoY; FY25 revenue up 46% over FY24.

• $124M+ in 2026 contracts, with 1,280 NVIDIA Blackwell GPUs and $21.9M prepaid by customers.

• On track to deploy ~6,900 GPUs by year-end 2026, up from 560 in June 2025.

• $90M convertible note facility and $26M equipment lease support capital needs.