Stock Taper Revenue: $6.7 million, up 118% year-over-year from $3.1 million.
Compute Revenue: $5.6 million, a 328% increase from $1.3 million a year ago, representing 84% of total revenue.
Gross Margin: Expanded to 66.6% from 55.1% year-over-year.
Operating Loss: Increased to $7.7 million from $2.2 million a year ago, primarily due to higher depreciation and amortization costs related to new HPC assets.
Net Loss: $22.8 million, compared to a net income of $12.1 million in the prior year, significantly impacted by non-cash accounting items.
Cash Position: Ended the period with $39.9 million in cash and restricted cash, up from $11.7 million at year-end.
GPU Fleet Growth: Increased from 952 to 3,088 GPUs, a 224% rise from Q1.
New Contracts: Signed 21 new direct customer contracts totaling $169.7 million, bringing total signed contract value to $282.5 million across 40 contracts.
Customer Base Shift: Direct customer relationships now account for over 96% of recurring revenue, up from less than 10% a year ago.
Colocation Agreement: Announced a new colocation agreement in Metro Atlanta for 3.75 MW, with potential for expansion.
Operational Focus: Emphasized a strategy of deploying GPU clusters rapidly in existing power locations rather than pursuing large-scale builds.
Capacity Target: Aiming for a total of 18 MW of HPC capacity by year-end, with 8 MW already leased and generating revenue.
Revenue Per Megawatt: Current contracts are generating $16 million to $20 million of annualized revenue per megawatt.
Demand Outlook: Not currently demand constrained, with a focus on acquiring additional power capacity to meet customer needs.
Operating Loss: The widening operating loss raises concerns about cost management as the company scales.
Net Loss: Significant net loss attributed to non-cash accounting items, which may obscure underlying operational performance.
Market Competition: Facing competitive pressures from larger players focusing on gigawatt-scale projects, while QumulusAI targets smaller, immediate needs.
Power Capacity Constraints: The primary limitation to growth is the availability of land and power for new deployments.
Customer Profiles: Focus on AI-native companies that require rapid compute capabilities, with plans to also target enterprise customers.
Financing Environment: The capital markets are evolving to support the fast-paced AI sector, with more financing options becoming available.
Guidance Reaffirmation: No reaffirmation of the previously stated $300 million ARR target, but management expressed confidence in meeting demand with current strategies.
Competitive Landscape: QumulusAI differentiates itself through speed and flexibility, catering to immediate customer needs rather than long-term, large-scale contracts. Overall, QumulusAI demonstrated strong growth in revenue and GPU deployment, with a strategic focus on rapid deployment and direct customer relationships, while also facing challenges related to operating losses and competitive pressures.
SOURCE: Q3 2026 EARNINGS CALL TRANSCRIPT