Reiner Pope of MatX on accelerating AI with transformer-optimized chips
Source
youtube.com
Author
Cheeky Pint
Date
Why it matters
Describes how accelerator design choices shape model speed and cost, and why new chips targeted at transformers may change inferenceRunning a trained model to get answers — the phase where AI is actually used, as opposed to trained.Full definition → economics.
Terms in this piece · Glossary
inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
