Design and Architecture — tvm 0.15.dev0 documentation
This document is intended for developers who want to understand the architecture of TVM and/or actively develop on the project. This page is organized as follows: The Example Compilation Flow gives an overview of the steps that TVM takes to turn a high level description of a model into a deployable module. To get started, please read this section first. The Logical Architecture Components section describes the logical components. The sections after are specific guides focused on each logical component, organized by the component’s name. The Device/Target Interactions page describes how TVM interacts with each supported physical device and code-generation target. Feel free to also check out the Developer How-To Guide for useful development tips. This guide provides a few complementary views of the architecture. First, we review a single end-to-end compilation flow and discuss the key data structures and the transformations. This runtime-based view focuses on the interactions of each com
Design and Architecture # This document is intended for developers who want to understand the architecture of Apache TVM and/or actively develop on the project. This page is organized as follows: The Overall Flow gives an overview of the steps that TVM takes to turn a high level description of a model into a deployable module. To get started, please read this section first. Brief introduction to the key components of the TVM stack. Feel free to also check out the TensorIR Deep Dive and Relax Deep Dive for more details about the two major components in the TVM stack. This guide provides a few c
Explore this link on the map →related reading
- A friendly introduction to machine learning compilers and optimizershuyenchip.com
- Tiny TPUtinytpu.com
- TPU Deep Divehenryhmko.github.io
- How To Scale Your Modeljax-ml.github.io
- PyTorch internals : ezyang's blogblog.ezyang.com
- UW PLSE | How does torch.compile work?uwplse.org
- Touching the Elephant - TPUs | Consider the Bulldogconsiderthebulldog.com
- NYSRGnotes.ekzhang.com
- How to Think About TPUs | How To Scale Your Modeljax-ml.github.io
- PiTorch: ML on Baremetal Raspberry Pis | projectsmasonjwang.com
- State of torch.compile for training (August 2025) : ezyang's blogblog.ezyang.com
- [1912.05036] RVSDG: An Intermediate Representation for Optimizing Compilersarxiv.org