HIVE Digital Technologies Ltd. (TSX: HIVE) (NASDAQ: HIVE) (BVC: HIVECO) said it has signed a five-year GPU cloud services agreement for about $350 million in total contract value, covering the delivery of a dedicated AI infrastructure cluster built around 2,016 NVIDIA Blackwell Ultra GPUs. The deal, announced on August 17, 2026 through HIVE’s wholly owned subsidiary BUZZ High Performance Computing Inc. (BUZZ HPC), adds approximately $70 million in annualized revenue and is expected to help BUZZ HPC’s GPU cloud platform reach full deployment by late 2026.
$350 million contract underpins BUZZ HPC’s long-term revenue run
The agreement is structured as a five-year commitment with an investment-grade enterprise customer. HIVE said the contract represents approximately $350 million in total contract value (TCV) over the term and is expected to add about $70 million of annualized revenue (ARR). Combined with BUZZ HPC’s current base, HIVE said BUZZ HPC’s total annualized revenue is approximately $180 million, made up of roughly $35 million of active, realized ARR today and about $145 million in contracted ARR expected to come online through Q4 2026.
HIVE also highlighted the cash-flow profile it expects from the deployment as the cluster scales to full operation. The company described the contract as providing “attractive long-term economics” and “meaningful recurring cash flow” once installed capacity reaches full deployment.
In describing the strategic rationale, HIVE said the five-year term is designed to provide longer-duration revenue visibility while supporting its approach to building accelerated computing infrastructure tied to Canadian “sovereign” AI requirements, in addition to serving global enterprises and AI organizations internationally. The company said BUZZ HPC will use its expertise across NVIDIA accelerated computing, high-performance networking, enterprise storage, and AI cloud infrastructure to support advanced AI training and inference workloads.
2,016 NVIDIA Blackwell Ultra GPUs targeted for Merritt AI Fabric
Under the contract, BUZZ HPC will deliver a dedicated AI infrastructure cluster featuring 2,016 NVIDIA Blackwell Ultra GPUs housed in NVIDIA GB300 NVL72 rack-scale systems. The cluster will be interconnected using NVIDIA Quantum-X800 InfiniBand networking and supported by VAST Data’s high-performance storage platform.
HIVE said the systems are being designed in accordance with NVIDIA’s reference architecture to support large-scale AI training, inference, and advanced enterprise workloads. The company also stated that the cluster is expected to become operational later this year at the Bell AI Fabric facility in Merritt, British Columbia.
The Merritt deployment is described as powered entirely by renewable hydroelectric energy. HIVE said it will also rely on closed-loop liquid cooling technology, which the company says eliminates ongoing water consumption for cooling—an element it linked to its efforts to deliver high-performance AI compute with a sustainability focus.
HIVE indicated that NVIDIA accelerated computing infrastructure is expected to be delivered and deployed in calendar Q4 of this year. At that point, the company said HIVE’s HPC/AI daily revenue is expected to be approximately $500,000.
Capital plan, deposit mechanics, and ownership of the hardware
To fund the deployment, HIVE said capital expenditures associated with the deployment—including the accelerated computing cluster and associated hardware and service warranties—are expected to total approximately $185 million.
HIVE also described the contract’s payment structure. It said the customer will provide an upfront deposit of approximately 10% of the total contract value. For the $350 million TCV, that implies a deposit of about $35 million. The company stated that this deposit, along with HIVE’s previously announced financing initiatives and additional equipment financing, is expected to support deployment and optimize its capital structure.
Crucially, HIVE said it will retain ownership of the NVIDIA AI infrastructure upon completion. The company characterized this as creating a long-lived strategic asset, with the possibility of continued cash flows over time. HIVE added that it expects the five-year contract to deliver an attractive return on invested capital, citing the long useful life of NVIDIA accelerated computing enabled by performance characteristics such as throughput per watt and “lowest token cost.”
HIVE’s management framed the investment and financing approach as part of a broader expansion plan for GPU-based AI infrastructure. It said the agreement is another step in scaling BUZZ HPC’s AI cloud platform and in securing large-scale, long-term commitments from investment-grade customers.
“A year ago, our team was focused on rapidly scaling HIVE’s ASIC computing power and Tier-I data center footprint by over 300%. Today, we are applying that same discipline and execution to scaling our GPU-based AI infrastructure.”
Frank Holmes, Executive Chairman of HIVE Digital Technologies, said he was “deeply pleased” with execution across HIVE and BUZZ teams and pointed to the company’s Canadian capacity plans, adding that with approximately 400 MW of capacity in Canada for Tier III data center development, HIVE believes it can bring more than 120,000 GPUs online over the next two years.
“We are pleased to announce this agreement as we accelerate towards our year-end target of $200 million of ARR for our GPU cloud business.”
Aydin Kilic, President and Chief Executive Officer of HIVE Digital Technologies, said the five-year term, expected stable cash flows from an investment-grade enterprise customer, and an upfront deposit of about 10% support an “attractive economic and return profile” for the deployment. Kilic also said that proceeds from HIVE’s June 2026 0% convertible bond will be used in the effort to bring the second long-term large GPU cluster deal forward, and that debt financing is being used to lever the purchase of GPUs.
Craig Tavares, President and Chief Operating Officer of BUZZ High Performance Computing, said the contract reflects BUZZ HPC’s ability to execute on large-scale AI infrastructure deployments while securing long-term, high-quality customers. He said BUZZ is establishing itself as a leading Canadian sovereign AI cloud provider, combining NVIDIA infrastructure with partners including Bell and VAST Data.
For more details, see the company’s announcement on Newsfile.
A renewable-power AI infrastructure push with enterprise cloud focus
HIVE described the Merritt facility as part of a broader AI cloud strategy that pairs computing performance with sustainable power infrastructure. The company said the deployment will operate on renewable hydroelectric power and use closed-loop direct liquid cooling technology, aimed at maximizing energy efficiency without ongoing water consumption for cooling.
Beyond the hardware delivery, the company said BUZZ HPC continues to scale its position as a Canadian AI cloud provider by delivering sovereign accelerated computing infrastructure and enterprise AI cloud services to customers in Canada and internationally. It said the agreement adds further evidence of customer demand translating into long-term contracted revenue.
In its statement, HIVE also reiterated the definition of ARR used in the release, describing it as run-rate revenue annualized based on realized revenue per week, per day, or per quarter, and noting that ARR projections may be unreliable indicators of future results due to possible cancellations, discounts, or service downgrades.
About HIVE Digital Technologies Ltd.
HIVE Digital Technologies Ltd. is a supplier of sustainable digital infrastructure, building and operating renewable energy-powered data centers used for high-performance AI computing. Through its wholly owned subsidiary BUZZ High Performance Computing Inc., HIVE delivers sovereign AI cloud infrastructure, with a dual-engine strategy that combines long-life digital infrastructure assets and renewable energy to generate diversified, long-term cash flows across AI infrastructure and digital asset mining.
