Bell Canada Expands AI Capacity with Bell AI Fabric – HostingJournalist.com

Bell Canada has unveiled a significant plan to enhance Canada’s AI infrastructure through the creation of a nationwide network of high-performance, hydroelectric-powered data centers called Bell AI Fabric. This initiative aims to offer up to 500 megawatts (MW) of computing capacity, potentially establishing the largest sovereign AI compute capability in the country.

This strategic move is designed to keep Canadian businesses competitive in the global AI landscape while bolstering national data sovereignty.

Bell AI Fabric is set to be a key element of Canada’s digital economy, providing advanced computing resources to businesses, researchers, and government bodies. The project also emphasizes Bell’s role as a comprehensive AI Managed Service Provider, supporting everything from infrastructure deployment to strategic AI application development. The initiative starts with a data center supercluster in British Columbia and is anticipated to expand quickly with multiple facilities already in various stages of planning and construction.

The first phase of Bell AI Fabric includes the deployment of two 7 MW AI inference centers, with the Kamloops, BC location expected to be operational by June 2025, followed by a second facility in Merritt by the end of the year. These initial centers will utilize Groq’s Language Processing Units (LPUs), designed to provide ultra-fast AI inference performance with lower operating costs, particularly advantageous for large language model workloads.

Bell’s partnership with Groq is crucial to the initiative’s technical foundation. Groq’s LPUs offer notable benefits over traditional processors, delivering faster inference speeds and significantly reduced cost per token. This integration, along with Bell’s existing fiber-optic infrastructure, aims to provide low-latency, scalable performance for AI applications across sectors like data centers, healthcare, and academia.

Mirko Bibic, President and CEO of Bell Canada and BCE, highlighted the long-term importance of the initiative. “Bell AI Fabric is transformative—not just for Bell, but for Canada’s AI economy. We are investing in a future where secure, sovereign, and sustainable AI computing capabilities are a national priority.”

Bell’s roadmap extends beyond the initial launch. In collaboration with Thompson Rivers University (TRU), a 26 MW AI data center will be built in Kamloops by 2026, followed by a second 26 MW facility in 2027. Both will be integrated with the BCNET network, providing high-speed connectivity and access to computing resources for TRU faculty and students. The design of the TRU data center will also include waste heat recovery to power campus buildings, reflecting a broader commitment to sustainable operations.

Further expansion includes two large-scale data centers currently in advanced planning that together will add over 400 MW of capacity. These will be fully powered by hydroelectric energy and engineered to support high-density AI workloads. Additional facilities will utilize Bell’s real estate assets across the country, enabling regional access to AI computing and reducing geographic bottlenecks.

Diana Gibson, British Columbia’s Minister of Jobs, Economic Development, and Innovation, commended the initiative’s potential to establish the province as a global AI hub. “British Columbia has a dynamic and growing AI sector. The Bell AI Fabric data center supercluster will not only drive technological innovation but also generate jobs and reinforce our status on the global stage.”

For the academic sector, the partnership between Bell and TRU offers an opportunity to democratize access to cutting-edge computing. Shannon Wagner, TRU’s interim provost, emphasized the educational impact: “This collaboration gives students and researchers access to high-performance AI resources that were previously out of reach. Kamloops will now be at the forefront of sustainable, applied AI innovation.”

Bell AI Fabric arrives at a time of rapidly increasing demand for AI infrastructure. With large-scale language models, advanced analytics, and machine learning workloads growing in complexity and volume, traditional infrastructure is struggling to keep pace. Bell’s focus on sovereignty, sustainability, and scale positions it to play a leading role in Canada’s AI future. As the network develops over the coming years, it could become the digital backbone for a new generation of intelligent applications and services rooted firmly on Canadian soil.