CB302B specs, module selection and setup: Documentation.

Technical guidance for teams evaluating, integrating, and deploying Curve Reality hardware.

Getting Started

Understand the platform and choose the hardware for your project.

Platform Overview

Architecture, software support, and deployment modes

Jetson Module Selection

Choosing an Orin Nano or Orin NX module for CB302B

Hardware Reference

Published specifications for the CB302B carrier board.

CB302B Datasheet

Hardware specifications in a printable web document

Board Specifications

Module support, display, power, and expansion interfaces

Software Support

Linux for Tegra (L4T) and native LineageOS 24 / Android 17 beta

Hardware Extras Setup

Install and check Curve's battery and sensor extras on a supported JetPack system.

Check Your JetPack Version

Find the installed JetPack and L4T release before you install

Install Hardware Extras

Supported configuration and installation command

Verify the Installation

Check battery status and sensor readings with curve-diag

Uninstall Hardware Extras

Remove the package and return to stock JetPack

Battery & Sensor Integration

Monitor the module with NVIDIA's tools, and understand the extras package and its published changes.

tegrastats, nvpmodel & jetson_clocks

Read tegrastats output and set power modes on Orin Nano and Orin NX

Battery Monitoring & Sensors

What the hardware extras package enables

Release Notes

Version history and fixes for hardware extras

05 - Memory architecture

Ember on Curve hardware.

Curve hardware pairs with Ember, our model-agnostic memory layer for persistent context across local and cloud models - continuity, recall, and personalization across device upgrades, model changes, and hybrid architectures.

Teams building persistent AI products use Curve hardware as the compute and power foundation, layering long-term memory behavior on top for more capable assistants, tools, and autonomous workflows.

Additional memory architecture patents are pending as Curve expands beyond hardware into infrastructure for persistent AI systems.

IP & patent position

06 - FAQ

Common questions.

Which Jetson modules are supported?

CB302B supports NVIDIA Orin Nano 4GB and 8GB and Orin NX 8GB and 16GB modules with Linux for Tegra (L4T) and native LineageOS 24 / Android 17 (beta).

Does the platform support battery-powered operation?

Yes. Curve Reality hardware is designed around portable, battery-capable deployment and supports hot-swappable M12 battery workflows for continuous operation.

Can the platform be used outside gaming and XR?

Yes. The same boards power robotics, embedded AI, industrial systems, and persistent AI product development - anywhere compute must be portable or battery-powered.

How does memory fit into the platform?

Curve hardware pairs with Ember, our model-agnostic memory layer for persistent context across local and cloud models. This gives teams continuity, recall, and personalization across device and model changes.

07 - Need more detail?

Talk to engineering.

Join early access or contact us directly for engineering support with platform integration.

Curve Reality - Rev 2026.04

Designed in New Zealand