BUZZ Deploys 2,016 Nvidia GPUs Under HIVE’s $350M AI Contract
HIVE Digital Technologies is deepening its push into artificial intelligence infrastructure, signing a five-year cloud services agreement that will deploy more than 2,000 Nvidia Blackwell Ultra GPUs in Canada.
The contract, secured through wholly owned subsidiary BUZZ High Performance Computing, carries a total value of about $350 million and is expected to contribute roughly $70 million in annualized revenue.
That would lift BUZZ HPC’s total annualized revenue to about $180 million. Roughly $35 million is already active, while another $145 million is contracted and expected to come online through the fourth quarter of 2026. The customer was described as an investment-grade enterprise but was not identified.
Nvidia Cluster Targets Large-Scale AI Workloads
BUZZ HPC will build a dedicated cluster using 2,016 Nvidia Blackwell Ultra GPUs in GB300 NVL72 systems. The deployment will also use Nvidia Quantum-X800 InfiniBand networking and VAST Data storage.
The infrastructure is designed to support large-scale AI training, inference, and other enterprise workloads.
The cluster is expected to begin operating later this year at Bell’s AI Fabric facility in Merritt, British Columbia. The site runs on renewable hydroelectric power and uses closed-loop liquid cooling, eliminating ongoing water consumption for cooling.
HIVE expects total capital expenditure for the deployment to reach about $185 million. The company plans to fund the remaining deployment through previously announced financing and additional equipment debt.
AI Revenue Becomes a Larger Part of HIVE’s Strategy
HIVE will retain ownership of the Nvidia infrastructure after the contract ends, giving it the potential to extend the assets’ revenue-generating life. Once the cluster is fully deployed, HIVE expects its HPC and AI business to generate about $500,000 in daily revenue.
CEO Aydin Kilic said the agreement moves the company closer to its year-end target of $200 million in GPU cloud annualized revenue.
“The five-year term, expected stable cash flows from an investment-grade enterprise customer, and upfront deposit of approximately 10% of total contract value support an attractive economic and return profile,” Kilic said.
Executive Chairman Frank Holmes framed the expansion as part of a broader “dual-engine strategy” spanning both ASIC-based digital infrastructure and GPU-based AI computing.
HIVE says it has about 400 megawatts of capacity in Canada available for Tier III data center development, potentially supporting more than 120,000 GPUs over the next two years.
The deal highlights how crypto infrastructure companies are increasingly monetizing their power and data center expertise through longer-term AI contracts, where revenue visibility can be far greater than in mining alone.
