MAX vs TVM

[hcye2022]

I have used TVM to deploy ML models on edge devices. It seems MAX has a lot of overlap with TVM. I am curious about whether MAX can overtake TVM in the long run, when more target platforms are added to Mojo --- would MAX match TVM features like Auto Tuning, what features can make MAX better than TVM, etc. --- general pros and cons of MAX vs TVM. Please chime in!

[Brian-M-J]

For those who don't know what TVM is: https://tvm.apache.org

In their LLVM talk, the team mentioned ML compilers in general. You can see what they said here.

Other than that, I don't have the necessary technical knowledge for any more comparisons.