TORONTO / CAMBRIDGE, Ontario — July 31, 2025. Canada Pension Plan Investment Board (CPP Investments) committed CAD 225 million—a 50% interest in the construction loan alongside Deutsche Bank Private Credit & Infrastructure—to fund a 54-megawatt (MW) hyperscale expansion at a data centre campus in Cambridge, Ontario. The project is pre-leased on a long-term basis to a GPU-focused AI cloud compute provider, anchoring capacity directly to machine-learning demand at go-live. Development is led by a joint venture of Related Digital, TowerBrook Capital Partners, and Ascent.

Why this matters beyond the server room: 54 MW at hyperscale is built for training and serving modern AI models and for high-throughput, real-time cloud applications. When that horsepower sits inside Ontario rather than far-flung regions, enterprises and end users see tangible benefits: lower latency, steadier throughput, and fewer time-outs during peak traffic. Those gains flow from industrial workloads to e-commerce front ends and into daily consumer apps. In that same stream of benefits, Ontario’s best online casinos—a large, regulated slice of the province’s digital economy—gain a sturdier foundation for smoother gameplay sessions, faster cashier flows, and more stable connections at peak load. (The policy backdrop is real: Ontario is actively prioritizing electricity access and approvals for data centres to support AI, cloud and digital services.)

What Was Announced: Financing, Partners, Tenant, and Market Signal

CPP Investments’ release details a CAD 225M commitment via construction financing (half the facility), with Deutsche Bank funding the other half. The Cambridge build adds 54 MW and is fully pre-leased to a market-leading AI cloud operator, reducing absorption risk and tying the site immediately to GPU-dense clusters once energized. CPP Investments framed the move as an extension of its global data-centre strategy in a Toronto-adjacent market where demand remains strong. A statement from Geoffrey Souter, Head of Real Assets Credit, cited the “rapid expansion of digital infrastructure” driven by cloud, data storage, and the “transformative potential of AI.” Trade coverage and local media echoed the core facts after the release.

Why Cambridge, why now. The Greater Toronto–Waterloo corridor concentrates enterprise HQs, universities, and cloud adoption, creating persistent pressure for near-user, low-latency capacity. Industry rundowns since July point to continued hyperscale absorption in the region; the Cambridge expansion fits that pattern by combining scale, a visible tenant, and proximity to Ontario’s core tech nodes.

Why 54 MW Changes AI Performance—and Everything Built on Top of It

AI workload fit

Fifty-four megawatts at hyperscale supports training runs for large models and high-availability inference for real-time services. Locating that capacity inside Ontario shortens round-trip times and reduces jitter between enterprise data sources, model endpoints, and end users—key for retail personalization, fraud detection, recommendation engines, and streaming/gaming session quality. CPP Investments

Policy tailwinds

Ontario has moved to unlock power for data centres and accelerate approvals where projects deliver economic benefits, explicitly linking these facilities to AI and cloud growth in the province. The government has said data centres are among the top new drivers of electricity demand in IESO’s outlook, and has publicly prioritized enabling capacity for the digital economy. That policy stance is the bridge between a single campus and a province-wide backbone.

Pipeline scale

Reporting in late July noted >6,500 MW of proposed Ontario data centres seeking grid access—illustrating both opportunity and the need to sequence power and approvals. The Cambridge financing is one of the near-term builds addressing that demand.

From AI to Checkout: The Cascade into E-commerce and Consumer Services

Step 1 — AI accelerates retail operations

With in-province compute, retailers can run heavier models closer to customers: e.g., real-time recommendations, dynamic pricing, vision-based quality checks, and same-session personalization without sending every call cross-border. Faster inference lowers drop-off at checkout; steadier throughput raises reliability during traffic spikes (seasonal sales, product drops). Nationally, retail e-commerce remains a meaningful and growing share of sales—$4.3B in Dec 2024 (6.1% of total retail)—and continued month-over-month retail growth in June 2025 underscores demand for robust digital infrastructure under the hood.

Step 2 — Cloud agility supports omnichannel

Hyperscale buildouts improve edge connectivity and interconnect options for payments, inventory, and last-mile APIs. That produces fewer checkout stalls, faster 3-D Secure challenges, and better fraud-model latency. Provincial think-tank and policy primers highlight AI’s economic potential for Ontario SMEs—provided compute is accessible and skills pipelines are built—another reason local capacity matters.

Step 3 — Online services benefit at peak

When compute sits closer, latency budgets shrink for telehealth, streaming, multiplayer gaming, and transaction-heavy services. Ontario’s government has explicitly tied data-centre policy to the digital services that residents use daily, from public to private providers.

Step 4 — A large regulated iGaming vertical rides the same rails

Ontario’s regulated market has scaled materially: FY 2024–25 revenue reached CA$3.20B, and July 2025 monthly revenue was reported at ~CA$311M (up ~29% YoY), with casino products dominating mix. The same low-latency, high-stability foundation that benefits retail and streaming also supports concurrency and live-dealer/video workflows in iGaming—one reason the sector’s performance is often cited as an indicator of consumer-facing digital maturity.

Regional Impact & Next Steps: Jobs, Interconnects, and a Stronger Backbone

Jobs and vendors

 Construction-loan projects of this size typically mobilize specialized trades (power, cooling, controls), carrier and dark-fiber partners, electrical OEMs, and long-lead equipment suppliers—economic spillovers that begin well before commissioning. CPP Investments’ release and subsequent trade coverage frame Cambridge as a near-market node that will add capacity quickly thanks to the pre-lease.

Connectivity flywheel

As new tenants land, operators add cross-connects and on-ramps that other enterprises can leverage, compounding the backbone effect. The province, for its part, has paired data-centre policy with broader digital-economy measures (e.g., compute access programs for SMEs) to push benefits beyond large cloud buyers.

Market context

Analysts and legal advisories summarizing IESO outlooks note that Ontario already leads Canada with 80+ data centres and expects data centres to represent ~13% of new electricity demand over the next decade—a scale that explains why targeted builds like Cambridge matter for the provincial grid and for digital competitiveness. Torys LLP

Cambridge’s 54-MW expansion isn’t just another build; it’s a practical way to keep AI training and inference close to Ontario’s users and enterprises. That proximity lifts everything stacked on top—retail e-commerce, payments, streaming and gaming, and the province’s large, regulated iGaming vertical—showing how targeted hyperscale investments translate into faster, smoother digital experiences across the province.