Baidu trained ERNIE 5.5 Flash Base on FlashMask-4 kernels, reporting speedups on NVIDIA SM100 GPUs of up to 175% in forward pass and 74% in backward pass over FlashAttention-4 MaskMod

441e85161Topic

AI infrastructure

A manufacturer-by-manufacturer overview of the chips, accelerators, and infrastructure powering modern AI systems.

2 debates·3,984 claims·624 news stories

0
Opening sign-in…