I created a repo that implements GEMM kernels for most datatypes supported on B200. It provides a python library that can also emit the `.cu` and `.cuh` files, so if you or an agent needs the raw files, the OpenGEMM library can provide the kernel without spending any tokens.
I created a repo that implements GEMM kernels for most datatypes supported on B200. It provides a python library that can also emit the `.cu` and `.cuh` files, so if you or an agent needs the raw files, the OpenGEMM library can provide the kernel without spending any tokens.
Blog: https://aramesh10.github.io/opengemm/opengemm.html