It's like a DigitalOcean dedicated to providing high-performance GPUs for AI applications.

Think of it as an express version of AWS for AI infrastructure.

• GPU-as-a-Service (GPUaaS): Provides rapid deployment of GPU-powered cloud infrastructure solutions tailored for artificial intelligence (AI) applications and high-performance computing (HPC) workloads.
• Blockchain Asset Management: Manages and operates infrastructure for blockchain assets, primarily focusing on grid power allocation for blockchain loads.

QumulusAI (QMLS) primarily sells its services to other companies and organizations.

Its major customer categories include:

• Small and mid-market customers, such as machine learning teams, AI infrastructure startups, and research institutions.
• Large enterprises, which utilize QumulusAI for long-term deployments or supplemental on-demand compute capacity.

NVIDIA (NVDA)

Michael Maniscalco, Chief Executive Officer, Chairman

Michael Maniscalco has served as QumulusAI's Chief Executive Officer and a member of its Board since September 2025. He was previously the Chief Technology Officer of Applied Digital Corporation (NASDAQ: APLD) from July 2023 to January 2025, where he transitioned the company's focus from blockchain data centers to high-performance computing (HPC) data centers and GPU-as-a-Service (GPUaaS). He also founded and successfully exited multiple ventures in chronic disease management, Internet of Things (IoT) and remote systems management, and intelligent building systems. Since February 2018, he has served as president of Pytheas Enterprises, a strategic consulting and fractional CXO services firm.

Scott Krosnowski, Chief Financial Officer

Scott Krosnowski has served as QumulusAI's Chief Financial Officer since February 2025. From October 2023 to February 2025, Mr. Krosnowski served as Senior Vice President, Finance, of private equity-backed go-to-market advisory firm, SBI Growth Advisory (Sales Benchmark Index, Inc.), where he supported M&A growth strategies. He was also the Chief Financial Officer of WAHA Technologies, Inc., a subsidiary of QumulusAI, from May 2022 to October 2023. Prior to that, from January 2016 to May 2022, he was the Chief Financial Officer of Arch Amenities Group (WTS International, Inc.), a global leisure management and consulting firm, where he was a minority owner and led infrastructure restructuring and supported rapid organic and M&A growth.

Ankur Chatterjee, Chief Integration Officer

Ankur Chatterjee has served as Chief Integration Officer since April 2023. He leads QumulusAI's HPC strategy and operations, including the deployment and management of GPU-intensive environments for AI infrastructure-as-a-service offerings. His career spans over two decades in technology infrastructure, systems integration, and operational leadership, including multiple ventures across private equity, cloud infrastructure, and edge computing.

Ryan DiRocco, Chief Technology Officer

Ryan DiRocco has served as Chief Technology Officer since September 2025. Prior to joining QumulusAI, he was the Chief Technology Officer at Performive, LLC, a VMware-focused managed multicloud provider, from June 2007 to October 2024. He brings over 20 years of expertise in IT, virtualization, enterprise networking, and managed services, having built and guided high-performing engineering teams.

Steve Gertz, Chief Growth Officer

Steve Gertz has served as Chief Growth Officer since December 2025 and previously as the Chairman of QumulusAI's Board from February 2025 to February 2026. He has been the founder of Rhythmic Ventures, a growth acceleration firm, since February 2017. Additionally, he was a co-founder of 11|TEN Innovation Partners, a healthcare-focused ecosystem and strategy consulting firm, from January 2017 to April 2024. He possesses over 25 years of experience in building high-impact partnerships across private equity, venture-backed companies, and large enterprises.

1. Execution Risk and Customer Concentration: QumulusAI faces significant execution challenges in rapidly scaling its High-Performance Computing (HPC) operations and transitioning its power capacity away from blockchain asset management. The company's growth strategy for 2026 relies on successfully deploying thousands of new GPUs and attracting sufficient customer demand to ensure HPC revenue surpasses blockchain revenue. A critical vulnerability is its high customer concentration, including an "85% dependency" on a single partner, RunPod, which poses a substantial risk if that partnership is disrupted. The success of these initiatives is subject to factors such as deployment timing, customer demand, power availability, and overall operational execution.
2. Capital Requirements and Funding for Infrastructure Expansion: The nature of the GPU-as-a-Service business is highly capital-intensive, demanding continuous and substantial investment in specialized hardware like NVIDIA GPUs and data center infrastructure. While QumulusAI has secured some capital through the sale of blockchain assets and other financing, its ambitious plans for future expansion, including acquiring more powered land and expanding its data center footprint, face uncertainty regarding the timing and scale of such expansions due to ongoing capital requirements and the availability of financing. Additionally, the availability and cost of power and GPUs are ongoing considerations that could impact expansion.
3. Competition and Rapid Technological Change: QumulusAI operates in the highly dynamic and competitive High-Performance Computing and Artificial Intelligence markets. Despite its focus on underserved small and mid-market segments, the company is exposed to intense competition from established large-scale cloud providers ("hyperscalers") and emerging specialized GPU-as-a-Service providers. The rapid pace of technological innovation in GPUs and AI architectures necessitates continuous adaptation and significant investment to maintain a competitive edge, with the inherent risk that existing infrastructure could quickly become less competitive or technologically obsolete.

QumulusAI (QMLS) is expected to drive future revenue growth over the next 2-3 years through several key initiatives:

1. Significant Expansion of High-Performance Computing (HPC) Operations: The company's growth strategy is focused on HPC, with an expectation that HPC compute services revenue will exceed blockchain operations revenue for the first time in 2026. This growth is underpinned by planned deployments of over 5,800 additional GPUs, increasing the total HPC IT load to approximately 11.0 MW by the end of 2026.
2. Accelerated Scaling of the HPC Business through Strategic Capital Allocation: QumulusAI received a $16.5 million cash injection from the sale of its interest in the T20 Mining Group joint venture. This capital is specifically intended to accelerate the development and scale of its HPC business. Additionally, the company plans to repurpose approximately 9 MW of on-grid power availability in Watonga for HPC data center expansion, representing approximately 6 MW of additional HPC IT load.
3. Growth in Underserved Small and Mid-Market Segments: QumulusAI specializes in providing flexible, competitively priced, and customizable GPU-powered solutions for AI applications, targeting small and mid-market customers, including machine learning teams, AI infrastructure startups, and research institutions. This focus on a market segment often overlooked by larger hyperscalers is expected to drive customer acquisition and revenue growth.
4. Rapid Deployment and Activation of GPU Infrastructure: The company distinguishes itself by deploying, activating, and producing revenue from GPU infrastructure within approximately 90 days, significantly faster than traditional providers who can take 12 to 24 months. This expedited path from procurement to revenue generation allows QumulusAI to scale its services and generate revenue more quickly from new capacity.

Inbound Investments

• In October 2025, QumulusAI secured a $500 million non-recourse financing facility, structured through blockchain and stablecoin liquidity channels, to fund GPU deployments.
• QumulusAI obtained $90 million in convertible notes from ATW Partners, consisting of two $45 million facilities in April and June 2026, with $30 million funded to date, specifically for NVIDIA Blackwell GPU procurement and data center infrastructure.
• A substantial minority investment was made into Global Digital Holdings, Inc., the predecessor to QumulusAI, in October 2023.

Outbound Investments

• Global Digital Holdings, Inc., which was later rebranded as QumulusAI, acquired The Cloud Minders, Inc. in April 2025, making it a wholly owned subsidiary. [BACKGROUND]

Share Issuance

• Between January 1, 2025, and June 26, 2026, QumulusAI issued shares to investors at prices ranging from $3.00 to $23.15 per share.
• The company completed a direct listing on the Nasdaq in July 2026, registering 37 to 39.5 million shares for resale by existing shareholders, but the company itself did not raise new capital through this listing.

Capital Expenditures

• QumulusAI reported capital expenditures of -$30.89 million in the last 12 months.
• The company plans significant expansion of its high-performance computing (HPC) business, with a 2026 roadmap that includes over 21,000 NVIDIA Blackwell GPUs planned across four quarters and an expectation to support approximately 11.0 MW of total HPC IT load by year-end 2026. [BACKGROUND, 6, 15]
• A $16.5 million cash injection from the sale of its interest in the T20 Mining Group, LLC joint venture in February 2026 is intended to accelerate the development and scale of its HPC business. [BACKGROUND] Additionally, QumulusAI entered into approximately $26 million in multi-year lease financing in May 2026 for a 50-node NVIDIA B200 GPU cluster.