HIVE Digital Technologies has signed a five-year, $350 million AI cloud contract that hinges on a $185 million infrastructure buildout before the majority of its projected revenue can be realized. The agreement was secured through HIVE's BUZZ High Performance Computing unit with an unnamed investment-grade enterprise customer.
Under the terms of the five-year agreement, BUZZ plans to install 2,016 NVIDIA Blackwell Ultra GPUs in GB300 NVL72 systems at HIVE’s Bell AI Fabric facility located in Merritt, British Columbia. The hardware, associated equipment, and service warranties are projected to cost approximately $185 million, with deployment targeted for the fourth calendar quarter of 2026.
The customer has agreed to provide an upfront deposit equivalent to roughly 10% of the contract value, amounting to about $35 million. HIVE has not yet disclosed whether this payment has been received, leaving the company reliant on existing financing and additional equipment funding to finalize the project.
To support GPU purchases, Chief Executive Officer Aydin Kilic pointed to bond and debt financing, noting that HIVE raised $130 million from zero-percent exchangeable senior notes in June and a total of $245 million from zero-percent notes during the quarter. However, the company has not disclosed the exact amount of those proceeds still available for the project, nor has it detailed the terms and closing status of additional equipment financing.
While the deal elevates BUZZ’s reported annual recurring revenue to approximately $180 million, only about $35 million of that total is currently active. An additional $145 million is contracted and anticipated to come online through the fourth quarter, meaning the bulk of HIVE's projected AI revenue depends heavily on hardware delivery, installation, and commissioning.
HIVE defines its annual recurring revenue using annualized weekly, daily, or quarterly figures, and cautioned that these projections may not account for future cancellations, discounts, or service reductions. Upon full operation of the new infrastructure, the company anticipates generating roughly $500,000 in daily revenue from its high-performance computing and AI business.
The company recently reported holding $208 million in cash, though it has not specified how much of those funds are designated for the GPU deployment. Consequently, the agreement establishes a substantial pipeline of committed AI demand, but its financial success relies on execution: securing remaining funding, deploying GPUs on schedule, and converting contracted capacity into active revenue.
