HIVE Digital Technologies Ltd. (TSX: HIVE) (NASDAQ: HIVE) (FSE: YO0) (BVC: HIVECO) said it has secured a major sovereign artificial intelligence infrastructure agreement through its wholly owned subsidiary BUZZ High Performance Computing Inc. (BUZZ HPC), partnering Bell Canada and Cohere Inc. The deal, described as a “full-stack” Canadian AI stack, centers on a three-year GPU cloud contract with a total contract value of approximately USD $220 million and will see the compute platform deployed at Bell’s facility in Merritt, British Columbia. The arrangement is intended to deliver production-grade sovereign AI infrastructure in Canada for enterprise and government customers, with the compute layer supporting Cohere’s foundation models and enterprise AI solutions.
A three-year GPU cloud contract for sovereign AI in Canada
Under the agreement, BUZZ HPC is executing a three-year GPU cloud contract valued at approximately USD $220 million. The company says HIVE will procure NVIDIA AI infrastructure as part of NVIDIA GB200 NVL72 rack-scale systems, powered by 2,304 NVIDIA Grace Blackwell GPUs. These systems will be interconnected using NVIDIA Quantum InfiniBand scale-out networking, configured to NVIDIA reference architecture standards and paired with advanced liquid cooling.
The infrastructure is set to be deployed at Bell’s purpose-built facility in Merritt, British Columbia. BUZZ HPC will provide the high-performance compute layer on which Cohere will operate its foundation models and enterprise AI solutions, targeting government and corporate customers across Canada. HIVE and its partners position the platform as a secure, national deployment designed to keep both compute infrastructure and model operations within Canadian borders.
HIVE said the deployment will be built on a unified Canadian infrastructure approach that combines Bell AI Fabric’s national data centre and connectivity platform, Cohere’s security-first enterprise-grade AI solutions and large language model capabilities, and BUZZ HPC’s NVIDIA-accelerated GPU cloud and “AI factory” expertise. The companies characterize the resulting stack as production-grade and sovereign, integrating connectivity, data centres, compute, professional services, and cybersecurity into a single national platform.
Security, renewable power and the architecture choices
The companies’ descriptions emphasize both security posture and operational design. BUZZ HPC’s AI cloud and GPU cluster infrastructure is expected to run within Canada under Canadian standards, with HIVE stating the compute infrastructure will remain entirely within Canadian borders. The press release also links the partnership to Canada’s federal AI strategy, which it says mandates investing in Canadian technology sovereignty and keeping Canadian data under Canadian control.
BUZZ HPC’s role is also framed around power efficiency. The firm says its AI factories are powered by renewable energy and are engineered for ultra-low power usage effectiveness (PUE), with industrial-scale GPU capacity positioned for one of Canada’s strategically important AI deployments. Beyond the GPU cluster and networking, the announcement also points to an “AI factories” model, stating that the compute environment is intended to support demanding AI workloads for training, fine-tuning and inference—though the press release ties these capabilities to Cohere’s foundation models and enterprise solutions in this specific deployment context.
Separately, the release describes BUZZ HPC’s overall managed services approach, noting support for Kubernetes, Slurm, virtual machines, and bare-metal deployments optimized for AI, machine learning, and scientific workloads. In the broader platform concept, this is presented as part of the operational governance and performance control sought by enterprise and government customers.
Financing, deployment timeline and the industrialization push
HIVE said it is funding the purchase of the NVIDIA Grace Blackwell rack-scale systems using a portion of the proceeds from its recent USD $115 million convertible note financing completed in April 2026. Hypertec, described as a Canadian 42-year OEM veteran in high-performance computing, is delivering a bespoke GPU server solution. Hypertec’s scope is listed as hardware procurement and system integration through to installation, commissioning, and OEM support.
In addition to the contract details, HIVE included executive commentary that highlighted both revenue targets and expected go-live timing for the NVIDIA GB200 deployment. The company said it is forecasting the deployment to go live in late 2026 to early 2027 and expects it to add approximately contracted USD $70 million in annual recurring revenue to its current realized USD $35 million ARR. Separately, HIVE said its contracted HPC revenue target has surpassed USD $100 million.
“Canada helped pioneer modern artificial intelligence. What we have lacked is not talent, it is industrial infrastructure to commercialize that talent at scale before others do it for us,” Frank Holmes, Executive Chairman of HIVE Digital Technologies, said in the release. “This partnership with Bell and Cohere is a defining moment. BUZZ HPC is the GPU factory layer that transforms Canada’s AI ambitions from political promises into productive national assets.”
HIVE also linked the deployment to a broader infrastructure thesis that frames AI scaling as dependent on factories and compute clusters. Aydin Kilic, President and CEO of HIVE, said: “We are very pleased to announce that our contracted HPC revenue target has surpassed $100m. We expect this NVIDIA GB200 deployment to go live in late 2026 to early 2027, adding approximately contracted $70m ARR to our current realized $35m ARR.”
Craig Tavares, President and COO of BUZZ HPC, positioned the deal as more than a single contract, describing it as the construction of an “AI manufacturing platform” for Canada and asserting that the deployment creates sovereign Canadian compute at industrial scale capable of running, fine-tuning, and serving advanced AI systems entirely within Canadian borders. The release attributes significance beyond the contract to an “AI rebuilding” dynamic in which countries that control infrastructure capture more economic value.
What the partnership combines across connectivity, compute and models
The companies described the partnership as strengthening BUZZ’s position in Canada’s sovereign AI cloud market alongside Bell AI Fabric’s role as the connective layer. In their framing, the combined platform integrates the components required to deliver AI services for enterprises and governments: connectivity via Bell’s national platform, data center capabilities, compute infrastructure from BUZZ HPC, professional services, and cybersecurity. Cohere’s enterprise AI solutions and large language model capabilities complete the stack, with the compute layer operating foundation models and supporting enterprise deployment needs.
Bell is identified in the release as Canada’s largest communications company, with emphasis on advanced fibre and wireless networks, enterprise services and digital media, and its use of cloud-based and AI-driven solutions. Cohere is described as a security-first enterprise AI company building foundation models and end-to-end products, offering an all-in-one platform designed for maximum security, privacy and deployment flexibility across clouds, private environments and on-premises. Hypertec’s involvement is tied to bespoke GPU server delivery and integration support.
For readers seeking additional details, the announcement is also available via source.
HIVE said the agreement represents a milestone in Canada’s AI infrastructure ecosystem, aiming to deliver production-grade sovereign AI infrastructure built in Canada for Canadian enterprise and government customers. The initiative combines Bell AI Fabric’s connectivity and data centre platform, Cohere’s security-first AI solutions and large language model capabilities, and BUZZ HPC’s NVIDIA-accelerated GPU cloud and AI factory capabilities into a unified national platform.
About HIVE
Founded in 2017, HIVE Digital Technologies Ltd. is a publicly listed company that initially prioritized mining digital assets powered by green energy. It now builds and operates Tier-I and Tier-III data centers across Canada, Sweden and Paraguay, serving Bitcoin and high-performance computing clients. HIVE’s infrastructure-driven approach includes hash rate services alongside GPU-accelerated AI computing.
