yeah even the official flashattention is moving many implementations from cutlass to triton except for the main mha backward/forward pass

It was written with cutlass? No wonder Peter Kim found it valuable and worthwhile to de-obfuscate. Adopting a new programming language invented by OpenAI doesn't sound like a much better alternative. I'd be shocked if either of them were able to build code for AMD GPUs, where it's easy to adapt CUDA code, but not if it's buried in tens of thousands of lines of frameworks. I like open source code to have clarity so I can optimize it for my own production environment myself. When people distribute code they've productionized for themselves, it squeezes out all the alpha and informational value. Just because something's open source doesn't mean it's open source. I think people mostly do it to lick the cookie without giving much away.

Triton has an AMD backend, although work is still ongoing.

You will also be able to use Triton to target Ryzen AI.

You mean triton the inference server or triton the DSL for cuda?

The DSL: https://openai.com/research/triton

they mean the dsl (not just necessarily for cuda)

triton the DSL.

> allows a level of abstraction that doesn't affect performance even a little bit

The second part of this sentence is true because the first part is false.

zero cost abstractions exist. doesn't mean all abstractions are zero-cost. or being zero-cost somehow invalidates their abstractness/genericness. but maybe we differ on the definition of abstractions.

As a person who finds CUDA extremely easy to write and integrate, what does Triton have to offer?

block level rather than thread level programming, automatic optimization across hyperparameters, makes it much easier to write fast kernels