AR Cloud
Overview
AR Cloud is a connected service that allows Magic Leap 2 devices to access large-scale spatial data. This spatial data can be managed on the web-based console where digital twin administrators create Shared Spaces, manage device-side scans, and merge scans to contribute to larger-scale Spaces. AR Cloud admins are able to import and export spatial maps, manage large-scale digital twin deployments, and monitor Magic Leap 2 device access.
See our guide on required licensing to be able to use AR Cloud on your device(s):
Benefits
AR Cloud allows users to go beyond the limitations of their Magic Leap 2 devices. The following features help enhance a user’s Magic Leap 2 experience:
- Co-present experiences – allows multiple users to experience the same digital content in the same place through shared maps and their spatial anchors.
- Large spatial maps (>10,000m2) – without AR Cloud, each device has a limit of five Local Spaces of ~250m2.
- Device access to an unlimited amount of Spaces – Magic Leap 2 devices can gain unlimited access to shared Spaces, as opposed to the 5 stored locally.
Getting Started
Choose Your Installation
AR Cloud can be installed to support both Virtual Machine installations and enterprise installations. Choose the correct installation for your needs. You can always change or update your installation later.
Criterion | Virtual machine | Cloud GCP, AWS, Azure | Custom private servers |
---|---|---|---|
audience | developers | enterprises | enterprises |
number of devices | ≤ 3 | ≥ 100 | ≥ 100 |
cost | low | medium-high depends on configuration | medium-high depends on configuration |
deployment target (skill-level) | virtualization platform (power user) | Kubernetes cluster (Kubernetes/cloud administrator) | Kubernetes cluster (Kubernetes/system administrator) |
production-ready scalable and fault-tolerant | ❌ | ✅ | ✅ |
supports big spaces | ❌ | ✅ | ✅ |
supports version upgrades preserves map data | ❌ | ✅ | ✅ |
supports private access | ✅ | ✅ / ❌ requires special configuration | ✅ |
Virtual Machine Installation
- Utilizes Virtual Machines and OTA/UTM files for an all in one installation.
- Runs on a desktop or laptop meeting the minimum specifications.
- Does not retain map data during upgrades of AR Cloud.
- Used to support a small number of devices (3 or less).
- Used by developers creating applications which utilize AR Cloud.
Enterprise Installation (Cloud, Custom)
- Runs on cloud providers such as AWS, GCP and Azure.
- Runs on custom servers meeting the minimum specifications.
- Fully scalable and customizable.
- Retains map data during upgrades of AR Cloud.
- Used to support a larger number of devices.
- Used by enterprises supporting production solutions.
System Requirements
Common Requirements
Minimum | Recommended | |
---|---|---|
Network bandwidth | 480 Mbps | 1 Gbps |
Wi-Fi version | Wi-Fi 5 (802.11ac) | Wi-Fi 6 (802.11ax) |
Storage | 100 GiB | 200 GiB |
Virtual Machine Requirements
Details can be found on the dedicated page.
Cluster Requirements (Cloud, Custom)
Minimum | Recommended | |
---|---|---|
Kubernetes version | 1.25.5+ | 1.27.3+ |
Node range | 2 - 6 | 4 - 12 |
Desired nodes | 4 | 8 |
vCPUs per node | 2 | 8 |
Memory per node (GiB) | 8 | 32 |
Each target environment provides examples of machines that can be used as nodes:
Compute platform optimizations: For optimal performance, it is suggested to use
microarchitectures that support x86-64-v3
. It is suggested to use x86-64-v3
nodes1 for the best performance.
- Compute nodes:
GCP: All example machine types supportx86-64-v3
. Microarchitectures used for specific machine types can be found here.
AWS: All example machine types supportx86-64-v3
. Microarchitectures used for specific machine types can be found here.
Azure: All example machine types supportx86-64-v3
. Microarchitectures used for specific machine types can be found here and here.↩
Used Components
For all enterprise installations, familiarity with Kubernetes is required to install AR Cloud.
The installation process also uses the following solutions, but detailed knowledge about them is not required:
Required third-party solutions deployed inside the Kubernetes cluster by default:
- PostgreSQL - DBMS
- MinIO - object storage
- NATS - message broker
- Keycloak - identity provider and auth server
Both the database and object storage can be replaced with external or managed services. Check the advanced setup guide for more information.
Leap Brush
After installation of AR Cloud, we welcome you to try our first-party, Open Source project, Leap Brush.
This project can be used as basis for building a multi-user, collaborative experience, leveraging AR Cloud's features. The project can work over local area network (LAN), enabling virtually no latency for your shared experience.
Support
For technical support, please reach out to us at care@magicleap.com.
Resources
📄️ GCP Deployment
Enterprise deployment to Google Cloud Platform (GCP)
📄️ AWS Deployment
Enterprise deployment to Amazon Web Services (AWS)
📄️ Azure Deployment
Enterprise deployment to Azure
📄️ Custom Deployment
Enterprise deployment to a custom server
📄️ VM Image Deployment
Basic deployment using a pre-configured virtual machine
📄️ Upgrade AR Cloud
Process for updating AR Cloud version after initial installation
📄️ Licenses
A breakdown of required licenses to operate AR Cloud
📄️ Enterprise Console
Access the Enterprise Console and manage devices
📄️ Customization and Security
Customize and secure your AR Cloud installation
🗃️ Integrations
1 items