Skip to main content

See HAMi at KubeCon China 2026 (Sept 7-9, Shanghai): keynote, sessions & booth T-1

HAMiHome Docs Tutorials Case Studies Community Events Blog Releases

Next
v2.9.0
v2.8.0
v2.7.0
v2.6.0
v2.5.1
v2.5.0
v2.4.1
v1.3.0

English
简体中文

Get started
Introduction
Install
Vendor guides
Operate
Design and develop
Contributor Guide

This is unreleased documentation for HAMi Next version.

For up-to-date documentation, see the latest version (v2.9.0).

Version: Next

Design and develop

Architecture, GPU internals, build and scheduler design.

📄️ GPU Virtualization

📄️ Architecture

📄️ HAMi Cluster Architecture After Installation

📄️ GPU Software Stack Overview

📄️ Understanding GPU Drivers

📄️ Ecosystem Integrations

📄️ Build HAMi

📄️ Protocol design

📄️ Scheduler Policy

📄️ HAMi-core design

📄️ Dynamic MIG Implementation

📄️ GPU Topology Scheduling

📄️ Kunlunxin Topology-Aware Scheduling

📄️ Profiling Scheduler

📄️ Scheduler Event Log

📄️ GPU Utilization Metrics

📄️ vNPU Core Integration

📄️ HAMi WebUI

📄️ HAMi mind map

📄️ Diagram Inventory

GPU Virtualization

Documentation

Concepts
Quick Start
Tutorials

Community

Discord
Slack (#hami-dev)

Follow Us

LinkedIn
X

© HAMi Authors 2026 | Documentation Distributed under CC-BY-4.0 | Powered by Netlify

Copyright © HAMi a Series of LF Projects, LLC
For website terms of use, trademark policy and other project policies please see lfprojects.org/policies/.

CNCF

HAMi is a CNCF Incubating project