A Vancouver startup just raised $100 million to fix a problem most AI conversations skip entirely. Wafr Technologies announced this week that it secured the funding from private investors to commercialize a cooling system that slashes water use at AI data centres by up to 95% and cuts cooling power consumption by up to 80%.
Founded in 2025 by Darrell Kopke and Bikram Singh, Wafr built a thermal battery that captures cooling capacity when grid electricity is cheap and releases it during peak demand. It's a smart piece of engineering aimed at a problem that's becoming impossible to ignore. A single 100-megawatt AI data centre in the U.S. consumes around 2 million litres of water per day, roughly the same as 6,500 households, according to a 2025 International Energy Agency analysis.
The Bottleneck Nobody Talks About
Everyone in AI talks about chips. Fewer people talk about the fact that those chips generate enormous heat, and cooling that heat down consumes staggering amounts of water and electricity. Kopke put it plainly: the AI infrastructure market has a major bottleneck, and it's power and water, not compute.
That's a useful reframe for business leaders evaluating AI vendors and infrastructure partners. I've sat across the table from executives who assume AI scaling is purely a chip and cloud-spend problem. It isn't. The physical constraints of running these systems, water availability, grid capacity, and cooling costs, are increasingly the limiting factor on how fast companies can deploy AI at scale.
Canada's Play for Infrastructure Relevance
Wafr plans to use the new capital to launch an AI research lab and data centre in Canada while hiring dozens of technical staff and expanding across the country. The company is targeting an additional $200 million and has already signed letters of intent with international partners building AI data centres, with the U.S. and Germany as priority markets. Kopke noted the U.S. has more than 5,300 data centres, while Germany, the next largest market, has roughly 400. That gap represents real commercial opportunity for a company solving the sustainability side of AI infrastructure.
Wafr's technology has already been demonstrated in India and Dubai, and Foresight Canada, which named the company one of its most investable cleantech startups in 2025, is backing the new research lab as a major contributor.
Why This Matters for Business
This fits a pattern I've watched develop over the past year. Canada isn't trying to out-compete the U.S. or China on foundation models. Instead, companies like Wafr are carving out a lane in the infrastructure layer, the unglamorous but increasingly critical work of making AI physically sustainable to run.
For any company evaluating cloud providers or data centre partners, sustainability credentials are becoming a genuine due diligence item, not a nice-to-have. Water and energy constraints are already triggering community pushback and regulatory scrutiny in multiple markets. Vendors who've solved for it will have a real advantage as scrutiny increases.
Looking Ahead
Wafr's $100 million raise is a strong vote of confidence, but the company still needs to prove its technology scales from demonstration sites to hyperscale facilities. If it works at scale, expect cooling efficiency to become a standard line item in how businesses evaluate AI infrastructure partners, right alongside price and uptime.
