HIVE Digital Technologies says its subsidiary BUZZ High Performance Computing has entered a strategic partnership with ProCogia, a Vancouver-based applied AI and data engineering company. ProCogia will procure dedicated GPU capacity from BUZZ HPC's Canadian data centers and provide software and implementation services on that infrastructure. [1]
HIVE Digital Technologies says its subsidiary BUZZ High Performance Computing has entered a strategic partnership with ProCogia, a Vancouver-based applied AI and data engineering company. ProCogia will procure dedicated GPU capacity from BUZZ HPC's Canadian data centers and provide software and implementation services on that infrastructure. [1]
This is an announced commercial partnership with a disclosed operating model. The public release does not give a contract value, a fixed term or a GPU quantity for the ProCogia arrangement. It should not be treated as a priced colocation lease or a newly opened data center.
ProCogia is positioned as both a compute customer and a services partner. It plans to use dedicated BUZZ capacity for its products, customer workloads and development environments. In the other direction, it becomes a preferred applied AI services partner for BUZZ customers that need model hosting, custom training, agents or integration work. [1]
The companies also intend to sell jointly into the United States and Europe. ProCogia's ZeroBoxx framework is expected to use BUZZ infrastructure as its primary cloud launch platform. The announcement names CallYeah, a healthcare voice AI product, and PolyKode, a code-migration product for regulated environments, as intended workloads. These are stated plans, not independently verified deployments or customer wins. [1][2]
The release places the relevant infrastructure in Canada. It does not assign ProCogia's capacity to a specific city, province or BUZZ facility. Vancouver identifies ProCogia's base, while the San Antonio dateline identifies the announcement's origin. Neither establishes where the contracted GPUs will run.
The United States and Europe are described as target customer markets. The release does not establish that this partnership adds data centers in either region. That distinction matters for a buyer whose requirement is tied to local latency or where information is processed.
For an enterprise trying to deploy AI, the buying decision can extend beyond leasing racks or reserving GPU hours. This partnership combines compute access with the work required to put a workload into production. The commercial question becomes who is responsible for infrastructure performance, application delivery and ongoing support.
A buyer should establish which entity signs its agreement, who invoices for compute and services, and how the parties handle incidents that cross both layers. It should also request the deployment location, capacity reservation terms, delivery date, data-handling commitments and exit provisions in writing. The announcement does not establish those customer-specific terms.
The companies describe the offering as sovereign Canadian AI infrastructure. That is their characterization. Buyers with residency or governance requirements still need to verify the treatment of application data, backups, logs, remote support and third-party services. A Canadian facility location alone does not answer every operational question.
The reviewed SEC exhibit and BUZZ's own release do not disclose the ProCogia commitment's dollar value, GPU model and count, MW requirement, duration, minimum spend or firm service-start date. We therefore assign no MW figure or rental price to this partnership. BUZZ's wider campus pipeline and separately announced customer contracts cannot be used as substitutes for this deal's undisclosed size. [1][2]
CleanSpark's September investor presentation identifies Anviran, LLC, a wholly owned Meta subsidiary, as the tenant for its Sandersville, Georgia, AI data center campus. Meta guarantees rent and operating expenses. The presentation describes 175 MW of contracted critical IT load, a 20-year base lease and approximately $6.6 billion of contracted lease payments over that base term. [1]
Read the analysisBroadcom's AI XPV financing platform connects Anthropic's compute expansion with capital from Apollo and Blackstone. Its subsequent SEC disclosure explains the five year lease backstop and potential liability behind the initial $35 billion tranche.
Read the analysis