Skip to main content
Back to top
Ctrl
+
K
TVM Homepage
Community
Download
Getting Started
Overview
Installing TVM
Install from PyPI
Install from Source
Docker Images
Quick Start
IRModule
TVM Errors
How To
Importing Models from ML Frameworks
End-to-End Optimize Model
Customize Optimization
Optimize Large Language Model
Cross Compilation and RPC
Export and Load Relax Executables
Mix Python/PyTorch with TVM Using BasePyModule
Bring Your Own Codegen
Deep Dive
Design and Architecture
Operator Fusion
Code Generation
External Library Dispatch (BYOC)
TVM Runtime System
Introduction to Module Serialization
Relax Virtual Machine
Pass Infrastructure
TVMScript
Device/Target Interactions
TensorIR
Tensor Program Abstraction
Understand TensorIR Abstraction
TensorIR Creation
Transformation
DLight: Rule-Based GPU Scheduling
MetaSchedule: Search-Based Auto-Tuning
Relax
Graph Abstraction for ML Models
Understand Relax Abstraction
Dataflow Pattern Language (DPL)
Relax Creation
Transformation
TIRx
Overview
Getting Started
Installation
Your first kernel
Programming Guide
CUDA Programming Guide
Defining a function
Data types and expressions
Buffers and memory
Control flow
CUDA C++/PTX intrinsics
Compiling and inspecting
Parser utilities
In-kernel profiling with CudaProfiler
Tensor Layout
Tile Primitives
Compiler Internals
TIRx lowering pipeline
Tile Primitive Dispatch
copy
copy → vec_auto global/shared path
copy → vec_auto register path
copy → ldstmatrix
copy → fallback
copy_async
copy_async → ldgsts
copy_async → tma_auto / tma_explicit
copy_async → dsmem
copy_async → smem->tmem (tcgen05.cp)
copy_async → tmem<->local (tcgen05.ld/st)
gemm
gemm_async
elementwise
elementwise → reg
elementwise → smem
reduction
reduction → local
reduction → shared
reduction → SM100 packed paths
permute_layout
Backend Internals
API Reference
Core TVMScript
Tile Primitive Authoring API
CUDA Authoring and Support APIs
Direct PTX Instructions
Trainium Authoring and Support APIs
Core TIRx IR
Layout IR
Execution Scope IR
Compilation
Compiler Transforms
Analysis
Tile Dispatch Extension API
Backend Extension API
API Reference
Python API
tvm.error
tvm.ir
tvm.arith
tvm.ir.instrument
tvm.ir.transform
tvm.target
tvm.driver
tvm.testing
tvm.exec
tvm.support
tvm.runtime
tvm.runtime.vm
tvm.runtime.disco
tvm.relax
tvm.relax.analysis
tvm.relax.backend
tvm.relax.block_builder
tvm.relax.distributed
tvm.relax.frontend
tvm.relax.op
tvm.relax.testing
tvm.relax.training
tvm.relax.transform
tvm.relax.dpl
tvm.s_tir.analysis
tvm.s_tir.schedule
tvm.s_tir.transform
tvm.s_tir.dlight
tvm.s_tir.backend
tvm.s_tir.tensor_intrin
tvm.script
tvm.script.parser
tvm.script.ir_builder
tvm.script.printer
tvm.te
tvm.topi
tvm.s_tir.meta_schedule
tvm.rpc
tvm.contrib
Other APIs
About
Contributor Guide
TVM Community Guidelines
Submit a Pull Request
Code Reviews
Committer Guide
Documentation
Code Guide and Tips
Testing TVM
Git Usage Tips
Using TVM’s CI
Release Process
Error Handling Guide
Publications
Security Guide
Index
Index
Repository
Show source
Suggest edit
Open issue
.rst
.pdf
Get Started
Get Started
#
IRModule
IRModule
Quick Start
Quick Start
