GPU
A computer chip built to do many calculations at once, used to train and run AI models.
A GPU, or graphics processing unit, was first designed to draw video game images. It is very good at doing thousands of simple calculations at the same time, which is exactly what AI models need. Most AI training and much inference runs on GPUs. They are expensive and often in short supply, so they drive AI costs.
Think of it as
Like a stadium of people each doing one simple sum at the same moment, instead of one expert doing all the sums one after another.
Example
A team's new model runs fine on an ordinary laptop for a demo. Serving a thousand customers at once needs rented GPUs in the cloud, which becomes their largest monthly bill.