Machine Learning
CUDA
A platform that lets AI models run on graphics chips (GPUs), training far faster than on a regular computer.
In real life
Tools that let AI run on graphics chips are a big reason training that once took weeks can take days.
Try it (2 minutes)
Time how long it takes one person to sort a deck of cards, then four people splitting it. Why is splitting the work faster?
Talk about it
Why would training a huge model need special, faster hardware?
In the game
Turn this card sideways to target System gets +1 Capability until end of turn. It costs 2 Data to play, and you can put it in your Training Set. How to play