Nvidia is buying CentML, a Toronto startup that helps AI models run far more efficiently, for over US $400 million. The price is a vote of confidence in the Canadian AI scene and a direct play at owning more of the AI stack.
What CentML Actually Does
CentML does not build models. It makes them faster and cheaper to run, the layer between raw GPU power and model performance. Its software finds smarter paths through compute traffic so workloads get done quicker.
Inside the Deal
Nvidia outbid multiple companies for CentML.
The final package? US $300M+ up front with bonuses tied to performance.
Almost the entire engineering team is coming over, including interns.
Toronto just became even more important in Nvidia’s global playbook.
Why This Hits Different for Canada
This is proof that deeptech coming out of Canadian labs has global pull.
The best AI minds are getting chances to stay and build at home, instead of heading to Silicon Valley.
Investors are already looking for the next CentML, and more funding could flow north. One exit does not make a pipeline, so watch whether the next founders build here or sell here.
What to Watch Next
Nvidia is likely folding CentML’s compiler tech into CUDA or Triton. That’ll push more AI workloads onto Nvidia’s ecosystem.
CentML’s optimizations could become a key part of how future LLMs and AI systems scale
More in the same thread on our Canadian AI companies reference.
