How have you seen Canada’s digital connectivity ecosystem change in your time in the industry, and what big forces are shaping the market today?
My journey began close to 30 years ago in telecommunications, working with equipment like Nortel and Nortel LTMS switching. Back then, picking up the phone and making a long-distance call connected through these systems. That world transitioned from brick buildings full of copper cabling and technicians into co-location and data centers. Early on, I helped establish a carrier-neutral facility at 151 Front Street in Toronto, one of Canada’s leading carrier hotels—at that time, probably 75% of all Canadian communications ran through that one building.
These carrier hotels are like hub airports. Once you are in, you can connect anywhere: Caribbean, another continent, or locally. They are the meeting spots for carriers, content providers, and consumers. Every market we enter—Toronto, Vancouver, Montreal—starts at these interconnection crossroads. On top of that, we have brought in enterprises needing low latency, then content providers, then cloud players like AWS and Google, and now AI stands on top of this foundation.
I see AI accelerating the data centre space in two key areas: learning (training algorithms) and inference (delivering results). Inference is still clunky—ask a deep query and you might as well make a sandwich while it loads—but the winners will be those who can deliver content instantly and affordably. Being in dense, carrier neutral data centers, such as Cologix, allows enterprises a strong position to accelerate how people consume information.
What makes Canada a unique market, and how do you see Cologix bridging the gap between the more connected and disconnected parts of the country?
Canada is unique because of its basics: cheap renewable energy, naturally cool climate—like a free air conditioner for data centers—stable regulation, and government strategies pushing digital infrastructure forward. Strong data sovereignty laws are also key. Businesses want information to stay local, and consumers now understand that when data goes to the cloud, it might actually sit in another country.
Sovereign cloud, sovereign AI, and sovereign connections put Canada in a very strong spot, ensuring data remains under Canadian jurisdiction.
Everybody claims security, but how do we know there will not be breaches, hackers, or misuse?
Breaches and hackers will always exist—we cannot eliminate that risk. What we can do is assume companies holding our data are protecting it as much as possible. AI will help us detect threats faster, but it is also arming bad actors with smarter tools. For me, sovereignty is key: if Canadian data stays in Canada, then even if there are issues abroad, our information is not open to another government’s retrieval. At least you are dealing with your own government, not one across the ocean.
The Canadian government has ambitious climate targets. Under your leadership, Cologix has GPUs as a service and almost 98% renewable energy. How do you achieve that and why does it matter?
Sustainability has become a non-negotiable. As an industry, power compression and GPU demand mean fossil fuels may play a short-term role, but at Cologix it is everything—we have invested more than $48 million since 2016 toward carbon neutrality and ESG goals. Our newest Montreal data center (MTL8) will use 100% hydroelectric energy from Hydro-Quebec, working toward LEED Gold certifcation, and span 205,000 square feet. Sustainability used to be a nice-to-have, now leading enterprises demand it from their partners.
Everyone is “drinking the juice” now, chasing ESG goals. Ten years ago, data centers were not seen as green—they needed megawatts of power and huge cooling systems. Today, efficiency means lower costs, better environmental outcomes, and stronger client appeal. Some clients choose our Montreal facilities specifically for 100% renewable energy, ranking it above price or reliability. So sustainability now directly drives business decisions.
You work with large as well as medium-sized companies. How fast is digital transformation happening across your portfolio of clients, and what examples stand out?
Our clients range from hyperscale players—like Google or Amazon—down to enterprises modernizing legacy IT stacks. My ideal customers are Canada’s biggest, oldest, and slowest companies, still running legacy servers and mainframes. They cannot move fully to the cloud yet, so they need hybrid solutions. Instead of spending $50–75 million on GPUs and IT staff, they can connect to GPU-as-a-service providers in our centers. This gives them low latency, reduced costs, and better security. For example, a decades-old retailer moving online can add AI chatbots, cloud storage, and efficiency without massive upfront capital or payroll shifts.
Can you explain ‘GPU as a service’ in layman’s terms? What could a client expect in terms of return on investment after three to six months?
Think of a call center, retailer, or bank adding an AI chatbot. Leadership is asking, “What is our AI strategy?” Instead of building GPU stacks, they can order services from providers in our data centers, paying monthly or yearly. A connection can be live in hours, saving capital and complexity. It is similar to the way cloud adoption worked: hybrid deployments, with some legacy equipment and some cloud services. Now we are seeing hybrid AI—businesses will control part of their IT deployment while using AI-as-a-service from providers next door.
With competition and edge computing on the rise, where is Cologix’s most exciting niche?
We are excited across the board, from scaling multi-megawatt deployments to expanding digital edge services for finance and e-commerce. But the most exciting niche is inference delivery for AI—providing results as close to the edge as possible, giving end users the best experience. Expansion will continue in Montreal, Toronto, and Vancouver, but also into secondary markets, bringing data, clients, and providers closer to the edge.
You have described many benefits—low latency, hybrid models, lower upfront costs. What would be the downsides?
A downside might be control. Some organizations, like governments or policing, may not want sensitive data with a third party. Outsourcing means relying on someone else—if their service fails, gets hacked, or prices rise, you are not fully in control. Long-term ROI might also favor building in-house. Overall, deployments in an interconnected data centre enables modernization.
