How to bring an ASIC Accelerator to TVM

At present, the AI ​​compiler we wrote ourselves uses the front end of relax, and then we have re-implemented the optimization, memory allocation, scheduling and code generation ourselves.

This is because we did not have the ability to understand the full-stack development of TVM at the beginning.

Our framework is becoming difficult to maintain recently, and I think it’s time to fully invest in TVM.

Any blogs, guides and other useful information are very valuable to me.

And I noticed How to Bring Your Own Codegen to TVM Is this feature still work?

I also want to know the BYOC status in Relax~~