Skip to main content
Ctrl+K
NVIDIA CUTLASS Documentation - Home NVIDIA CUTLASS Documentation - Home

NVIDIA CUTLASS Documentation

NVIDIA CUTLASS Documentation - Home NVIDIA CUTLASS Documentation - Home

NVIDIA CUTLASS Documentation

Table of Contents

  • Changelog

CuTe DSL

  • Overview
  • Functionality
  • Quick Start Guide
  • CuTe DSL
    • Introduction
    • Code Generation
    • Control Flow
    • JIT Argument Generation
    • JIT Argument: Layouts
    • Struct-like JIT Arguments
    • JIT Caching
    • JIT Compilation Options
    • JIT Types
    • Integration with Frameworks
    • Debugging with the DSL
    • IKET Profiling
    • Autotuning with the DSL
    • Educational Notebooks
    • Deprecation Policy
    • Compile with TVM FFI
    • Ahead-of-Time (AOT) Compilation
    • Talks and Presentations
    • Naming Conventions
    • MMA Programming Guides
      • Warp-Level MMA Instructions Programming Guide
      • Warpgroup MMA Programming Guide
      • tcgen05 MMA Programming Guide
  • CuTe DSL API
    • changelog
    • cute
      • arch
      • Runtime
    • cute_nvgpu
      • Common
      • warp submodule
      • warpgroup submodule
      • cpasync submodule
      • tcgen05 submodule
    • pipeline
    • utils
      • Utilities for SM90
      • Utilities for SM100
  • Limitations
  • FAQs

CUTLASS Operator API

  • CUTLASS Operator API
  • Tutorials
    • Core concepts and basic GEMM
    • GEMM with fused epilogue
    • Bringing your own kernel
    • Host latency best practices
    • Fake tensors for compilation
    • Grouped GEMM with contiguous offset
    • Block-scaled GEMM (MXFP8)
  • API reference
    • Operator interface
    • Arguments and Operands
    • Kernel discovery
    • Metadata
    • Miscellaneous

CUTLASS C++

  • Overview
  • Getting Started
    • Quickstart
    • IDE Setup
    • Build
      • Building on Windows with Visual Studio
      • Building with Clang as host compiler
    • Functionality
    • Terminology
    • Fundamental Types
    • Programming Guidelines
    • GEMM Heuristics
  • Efficient GEMM in CUDA
  • Synchronization primitives
  • CUTLASS Profiler
  • Dependent Kernel Launch
  • Blackwell Specific
    • Blackwell SM100 GEMMs
    • Blackwell Cluster Launch Control
  • CuTe
    • 00_quickstart
    • 01_layout
    • 02_layout_algebra
    • 03_tensor
    • 04_algorithms
    • 0t_mma_atom
    • 0x_gemm_tutorial
    • 0y_predication
    • 0z_tma_tensors
  • CUTLASS 3.x
    • Design
    • GEMM Backwards Compatibility
    • GEMM API
  • CUTLASS 2.x
    • Layouts and Tensors
    • GEMM API
    • Tile Iterator Concepts
    • Utilities
  • Code Organization
  • Grouped Kernel Schedulers
  • CUTLASS Convolution

Reference

  • Software License Agreement
  • API reference

API reference#

API Reference for the public interface of cutlass.operators.

For a more step-by-step introduction to core concepts and worked examples, see Tutorials.

The API reference is grouped into the following pages:

  • Operator interface
  • Arguments and Operands
  • Kernel discovery
  • Metadata
  • Miscellaneous

previous

Block-scaled GEMM using CUTLASS Operator API

next

Operator interface

NVIDIA NVIDIA

Copyright © 2025-2026, NVIDIA Corporation.

Last updated on Jul 08, 2026.