Introduction
This document describes how to compile and install TVM 0.9.0 and optimize the performance in the openEuler 22.03 LTS SP3 operating system (OS) based on the new Kunpeng 920 processor.
Overview
Tensor Virtual Machine (TVM) is an open source deep learning compiler stack developed by the University of Washington and Allen Institute of Artificial Intelligence (AI2). It aims to optimize and efficiently deploy machine learning models on various hardware backends. TVM automatically optimizes computational graphs and operators to support high-performance inference across hardware such as CPUs, GPUs, NPUs, and FPGAs. It is widely used in edge computing, inference running on servers, AI chip adaptation, and more scenarios.
For more information, visit the TVM official website.
Programming language: C++ and Python
Brief description: a cross-hardware compiler optimized for deep learning models
Open source license: Apache License 2.0
When using open source software, comply with the applicable license.
Recommended Software Version
TVM 0.9.0