Glossary

The words to understand what comes next.

Clear definitions of the technical, financial, and policy terms used in Canadian tech.

Glossary results

Compute

IPU

A specialized processor that handles machine-intelligence workloads efficiently.

Definition and context : IPU
Definition
IPU refers to an intelligence processing unit. It is a design label for processors that focus on machine-learning workloads rather than broad consumer computing.
Why it matters
Readers should treat IPU coverage as a signal that the compute stack is being reshaped around inference, orchestration, and lower-latency AI delivery.
Aliases
intelligence processing unit
Related topics
Inference · AI acceleration · Model deployment
Compute

GPU

The general workhorse chip for training and running modern AI systems.

Definition and context : GPU
Definition
GPU stands for graphics processing unit. It was originally built for graphics, but its ability to process many operations in parallel makes it a core part of AI training and inference.
Why it matters
GPU availability often sets the pace for AI deployment, pricing, and product roadmap decisions.
Aliases
graphics processing unit
Related topics
AI infrastructure · Inference · Cloud compute