Re:Demystifying GPU Compute Architectures – By Babbage
by cavisne on Tuesday February 06, 2024 @09:57PM
(#39283918)
You don’t have to go to that level of detail but it is still relevant. For example FlashAttention is built on optimizing for this level of detail.
[ Parent
| on HN ]
Re:Demystifying GPU Compute Architectures – By Babbage
by TechnicolorByte on Wednesday February 07, 2024 @02:50AM
(#39285741)
Any specifics you can describe or somewhere to read about this?
Re:Demystifying GPU Compute Architectures – By Babbage
by WithinReason on Wednesday February 07, 2024 @04:00AM
(#39286176)
Just google Flash Attention 2, first hit is the paper.