Getting Started

• Quickstart
  • Prerequisites
  • Initial build steps
  • Build and run the CUTLASS Profiler
  • Build and run CUTLASS Unit Tests
  • Building for Multiple Architectures
  • Using CUTLASS within other applications
  • Launching a GEMM kernel in CUDA
  • Launching a GEMM kernel using CUTLASS 3.0 or newer
  • CUTLASS Library
  • Example CMake Commands
  • GEMM CMake Examples
  • Convolution CMake Examples
  • Instantiating a Blackwell SM100 GEMM kernel
• IDE Setup
  • Overview
  • VSCode Setup
  • clangd Setup
• Build
  • Building on Windows with Visual Studio
  • Building with Clang as host compiler
• Functionality
  • Device-level GEMM
  • Device-level Implicit GEMM convolution
  • Warp-level Matrix Multiply with Tensor Cores
  • Warp-level Matrix Multiply with CUDA WMMA API
• Terminology
• Fundamental Types
  • Numeric Types
  • Containers
  • Functional
  • Numeric Conversion
• Programming Guidelines
  • Hierarchical Organization
  • Design Patterns
  • Style
  • CUTLASS idioms