Skip to content

100 days of learning & making kernels in cuda / triton

Notifications You must be signed in to change notification settings

SwekeR-463/100kernels

Repository files navigation

100 Days of Kernels


  • Day - 1 -> Vector Addition Kernel

  • Day - 2 -> Vector Multiplication Kernel

  • Day - 3 -> Matrix Transpose Kernel & RGB to GreyScale Kernel

  • Day - 4 -> Matrix Addition Kernel

  • Day - 5 -> Matrix Multiplication Kernel


  • Day - 7 -> Lightning Attention Kernel

  • Day - 8 -> Tiled Matrix Multiplication Kernel





  • Day - 13 -> Basic 1D Kernel & Tiled Cached 1D Kernel








  • Day - 21 -> Binary Cross Entropy Loss Kernel


  • Day - 23 -> ConvNet Forward & Backward Kernel

  • Day - 24 -> cuBLAS dot product Kernel & cuBLAS convnet forward pass Kernel


  • Day - 26 -> cuBLAS Strided Batched GEMM Kernel & Naive Layer Norm Kernel

  • Day - 27 -> Optimized Layer Norm Kernel


  • Day - 28 -> Tensor Core Mat Mul Kernel & Optimized RMS Norm Kernel

About

100 days of learning & making kernels in cuda / triton

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published