Alauda Build of Hami

Introduction

TOC

What is HAMi?

Heterogeneous AI Computing Virtualization Middleware (HAMi), formerly known as k8s-vGPU-scheduler, is an "all-in-one" chart designed to manage Heterogeneous AI Computing Devices in a k8s cluster. It can provide the ability to share Heterogeneous AI devices among tasks.

Why HAMi

Device sharing

  • Support multiple Heterogeneous AI Computing devices
  • Support device-sharing for multi-device containers

Device Memory Control

  • Hard limit inside container
  • Support dynamic device memory allocation
  • Support memory allocation by MB or by percentage

Device Specification

  • Support specify a type of certain heterogeneous AI computing devices
  • Support specify a certain heterogeneous AI computing devices using device UUID

Easy to try

  • Transparent to tasks inside container
  • Install/Uninstall using helm, easy and green

Open and Neutral

  • Jointly initiated by Internet, finance, manufacturing, cloud providers, etc.
  • Target for open governance with CNCF