skills.rest
NVIDIA-AI-IOT
NVIDIA-AI-IOTOFFICIAL
Explore all skills by this author →
WebsiteGitHub

jetson-optimize-memory

Disable unused subsystems to reclaim DRAM in Jetson deployments.

46|7|Updated Apr 27, 2026
Search Tags:#memory optimization#jetson#DRAM carveout#headless deployment#SWIOTLB
One-click install
npx skills add https://github.com/NVIDIA-AI-IOT/jetson-bsp-skills --skill jetson-optimize-memory-nvidia-ai-iot
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: jetson-optimize-memory
Source: https://github.com/NVIDIA-AI-IOT/jetson-bsp-skills/tree/main/skills/jetson-optimize-memory
Command: npx skills add https://github.com/NVIDIA-AI-IOT/jetson-bsp-skills --skill jetson-optimize-memory-nvidia-ai-iot

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps reclaim DRAM by disabling unused subsystems across various layers of MB1 BCT, MB2 BCT, kernel reserved-memory, and SWIOTLB for headless or no-camera Jetson deployments.

Core Features & Use Cases

  • Memory Optimization: Reclaim DRAM by disabling unused subsystems.
  • Headless and No-Camera Configurations: Designed for scenarios where the deployment does not require display or camera.
  • SWIOTLB Management: Provides an option to shrink SWIOTLB DMA bounce pool for peripherals.

Quick Start

Reclaim unused memory in your Jetson device with /jetson-optimize-memory headless | no-camera.

Frequently Asked Questions about jetson-optimize-memory

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I reclaim DRAM on Jetson for headless deployments?▼

Reclaim DRAM on Jetson by disabling unused subsystems across MB1 BCT, MB2 BCT, and kernel reserved-memory settings. This optimization is specifically designed for headless and no-camera deployments where display and camera subsystems are unnecessary.

What is SWIOTLB management and how does it optimize Jetson memory?▼

SWIOTLB management shrinks the DMA bounce pool for peripherals to free up DRAM. By reducing the SWIOTLB allocation on Jetson devices, you reclaim additional memory that would otherwise remain reserved for peripheral DMA operations.

Can I use Jetson memory optimization for camera-enabled deployments?▼

Jetson memory optimization targets headless and no-camera configurations by disabling unused display and camera subsystems. Camera-enabled deployments require these subsystems active, making this memory reclamation approach unsuitable for that scenario.

What firmware settings are required for Jetson DRAM carveout?▼

Jetson DRAM carveout requires adjustments in MB1 and MB2 BCT firmware settings alongside kernel reserved-memory configurations. Specific carveouts vary depending on the chip architecture of your Jetson device.

When should I not disable unused subsystems for Jetson memory optimization?▼

Avoid disabling unused subsystems when your Jetson deployment requires display output or camera functionality. The memory reclamation strategy specifically targets scenarios where these subsystems remain unused throughout the device operation.

Related Skills

Explore More Software Engineering Skills→
JORDAN0615JORDAN0615

jetson-optimize-memory

Reclaim unused DRAM by disabling peripheral carveouts across Jetson boot layers.

Community
Advanced
NVIDIA-AI-IOTNVIDIA-AI-IOT
110

jetson-memory-audit

Measure Jetson DRAM and NvMap usage on live NVIDIA Jetson devices.

Official
Advanced
Marwan78888Marwan78888
16

embedded-systems

Automates embedded firmware design and optimization workflows for STM32 and ESP32 microcontrollers.

Community
Advanced
NVIDIA-AI-IOTNVIDIA-AI-IOT
110

jetson-speculative-decoding

Configure speculative decoding for Jetson vLLM servers with EAGLE-3 heads or draft models.

Official
Intermediate