Examples#

• Basic GEMM

  • Declaring and running a GEMM
  • Changing operation modes
  • Running cached kernels
  • Running non-default GEMMs
  • Handling errors

• Epilogue

  • Run a GEMM with an identity activation function
  • Run a GEMM with a ReLU element-wise activation function
  • Other element-wise activation functions

• PyTorch Extension

  • Background on grouped GEMM
  • Declaring a grouped GEMM via the CUTLASS Python interface
  • Exporting the CUTLASS kernel to a PyTorch CUDA extension