Reiner Pope of MatX on accelerating AI with transformer-optimized chips
6:26a tautological answer, basically. Is there some way you can have a mental
model for why that is the case?19:05build out,
there is dependency on logic ties from typically TSMC, maybe Samsung, or
HBM, which are the big three HBM vendors.21:35some are fungible.
Logic ties are typically pretty fungible. But other parts of manufacturing,
you actually need something














