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

jetson-memory-audit

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

110|19|Updated May 25, 2026
Search Tags:#audit#memory#diagnostics#jetson#nvmap#reclamation
One-click install
npx skills add https://github.com/NVIDIA-AI-IOT/jetson-device-skills --skill jetson-memory-audit
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: jetson-memory-audit
Source: https://github.com/NVIDIA-AI-IOT/jetson-device-skills/tree/main/skills/jetson-memory-audit
Command: npx skills add https://github.com/NVIDIA-AI-IOT/jetson-device-skills --skill jetson-memory-audit

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jetson-diagnostic, jq, python3, sudo, sysctl, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps you accurately measure live memory usage on NVIDIA Jetson devices and verify whether memory actually freed after a workload change, instead of relying on guesses or stale readings.

Core Features & Use Cases

  • Live Memory Snapshot: Captures Jetson DRAM usage, available memory, top consuming processes, and NvMap attribution when available.
  • Before/After Reclamation Checks: Compares snapshots around actions like stopping a model server, disabling the desktop, or changing runtime settings.
  • Stuck Memory Guidance: Explains when cache dropping is appropriate on affected JetPack and L4T releases and provides a safe host-side verification workflow.

Quick Start

Ask the assistant to run the jetson-memory-audit skill and return a current memory snapshot for the Jetson so you can see what is using memory right now.

Frequently Asked Questions about jetson-memory-audit

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

FAQPage Schema
How do I check Jetson memory usage and verify reclamation after stopping a GPU workload?▼

To verify Jetson memory reclamation, capture a baseline DRAM snapshot before stopping a GPU workload, then take an after snapshot to compare available memory and NvMap attribution differences.

What is NvMap attribution and how does it relate to Jetson DRAM diagnostics?▼

NvMap attribution tracks memory allocated through the NVIDIA Jetson NvMap driver. It provides granular DRAM diagnostics by attributing memory usage to specific processes and GPU workloads.

Does the jetson-memory-audit skill work on headless Jetson devices?▼

Yes, the audit supports baseline capture before switching to headless-mode. It measures live DRAM usage and verifies memory reclamation on headless Jetson devices using host-visible memory data.

Why does Jetson memory not free up after stopping a model server, and can I drop cache safely?▼

Memory may remain allocated due to NvMap heap retention or system cache. The skill provides stuck memory guidance and a safe workflow with authorized cache flushing to verify actual reclamation.

Do I need sudo or specific dependencies to run a Jetson memory audit?▼

Yes, running a Jetson memory audit requires sudo, python3, jq, and jetson-diagnostic. These dependencies enable authorized cache flushing, process attribution, and safe snapshot comparisons.

Related Skills

Explore More Software Engineering Skills→
NVIDIA-AI-IOTNVIDIA-AI-IOT
110

jetson-diagnostic

Capture a read-only Jetson health snapshot and memory summary.

Official
Advanced
NVIDIA-AI-IOTNVIDIA-AI-IOT
110

jetson-inference-mem-tune

Selects Jetson inference runtime and memory flags from an audit JSON for LLM/VLM serving workloads.

Official
Advanced
JORDAN0615JORDAN0615

jetson-optimize-memory

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

Community
Advanced
jralphjralph

performance-go

Profile Go applications with pprof to identify and resolve performance bottlenecks.

Community
Advanced