LiveFR

GPU

Definition: A GPU is a processor built for massively parallel computation, making it the reference hardware to train and run AI models.

Neural network operations are largely matrix multiplications, which GPUs execute very efficiently. Their availability is a key driver of AI cost and speed.

See also

← Full AI glossary · AI news