# BCE Secures Compute Deal to Host Cohere Models > Source: > Published: 2026-06-18 15:53:37.583160+00:00 # BCE Secures Compute Deal to Host Cohere Models According to a PR Newswire release and contemporaneous reporting by Bloomberg and Yahoo Finance Canada, Bell Canada (BCE) has announced a collaboration with **Cohere**, **Hypertec**, and **BUZZ High Performance Computing (BUZZ HPC)** to host and operate AI models on Bell AI Fabric infrastructure. PR Newswire states that compute capacity will come from Bell's data centre in **Merritt, British Columbia**, and that BUZZ HPC will deliver an AI-native cloud layer using Quebec-based **Hypertec** hardware and NVIDIA's **DSX AI factory** platform. Bloomberg frames the agreement as part of BCE's ongoing push to expand its AI strategy, and PR Newswire links the deal to the Canadian Sovereign AI Alliance initiative. Michel Richer, president of Bell AI Fabric, is quoted in the release on the partnership's role in moving workloads to production on Canada-located infrastructure. ### What happened According to a PR Newswire announcement, **Bell Canada (BCE)**, **Cohere**, **Hypertec**, and **BUZZ High Performance Computing (BUZZ HPC)** announced a multi-party agreement to host and operate artificial intelligence models on Bell AI Fabric infrastructure. The release states that compute capacity will come from Bell's data centre in **Merritt, British Columbia** and that BUZZ HPC, a subsidiary of Hive Digital Technologies, will provide an AI-native cloud layer. PR Newswire says the platform will use hardware manufactured in Canada by **Hypertec** and NVIDIA's **DSX AI factory** platform to support accelerated, production-grade AI workloads. Bloomberg and Yahoo Finance Canada report the deal is framed as continuing BCE's effort to expand its AI offerings and to support Canadian-sovereign infrastructure. ### Technical details Per the PR Newswire release, the collaboration combines Bell AI Fabric's data centre and connectivity, Cohere's enterprise-grade AI models, BUZZ HPC's scalable accelerated computing layer, and Hypertec-supplied hardware. The release describes the stack as built to enable R&D and production deployment for foundation models on Canadian infrastructure and notes the involvement of NVIDIA DSX as the underlying accelerated computing platform. ### Editorial analysis Companies building sovereign-AI stacks typically assemble a telecommunications or data-centre owner, an AI model vendor, and an accelerated-compute integrator to address latency, data residency, and procurement friction. This agreement follows that pattern: a carrier-owned data centre provides colocated capacity, a model provider supplies foundation models and enterprise tooling, and a specialized HPC integrator layers in GPU acceleration and system orchestration. ### Context and significance For practitioners, the deal matters because it reflects a growing market demand for locally hosted, enterprise-grade AI infrastructure in regulated jurisdictions. Having an integrated offering that couples local data centres with GPU-accelerated nodes and an enterprise model vendor reduces some operational friction for customers constrained by data residency or procurement rules. It also highlights continuing commercialisation efforts by model vendors like Cohere to serve government and enterprise clients via hosted, managed deployments. ### What to watch Observers should track: - •service-level details and pricing when they are published - •the types of model deployments and governance controls Cohere offers on the platform - •interoperability with customers' on-premises systems and common MLOps tooling. Also watch for capacity rollouts at the Merritt facility and any benchmarked performance figures for NVIDIA-accelerated workloads that BUZZ HPC or Hypertec publish ### Quote from the announcement "This agreement underscores the role Bell AI Fabric is playing in helping organizations move from experimentation to production on infrastructure that is located, operated and governed in Canada," said Michel Richer, president of Bell AI Fabric, in the PR Newswire release. ### Limits of reporting The public release describes the collaboration and the intended technical components; it does not include detailed pricing, specific SLAs, or a public roadmap for roll-out timelines. PR Newswire includes the customary forward-looking statements disclaimer in the release. ### For practitioners Enterprises and system integrators evaluating sovereign deployments should treat this announcement as a signal that integrated, carrier-backed AI hosting offers are becoming commercially available in Canada, and should plan procurement and compliance evaluations accordingly. ## Scoring Rationale This is a notable infrastructure deal that expands sovereign, production-grade AI hosting options in Canada. It is directly relevant to enterprises and MLOps teams but does not introduce new model capabilities or global-scale market disruption. Practice interview problems based on real data 1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)