NVIDIA Corporation

Collects a read-only health snapshot of a Jetson device, covering its model, memory, GPU, temperatures, power mode, storage, services, and processes. Helps investigate slowdowns, overheating, and memory pressure using observed device data.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Identify the Jetson model and memory capacity
• Investigate slowdowns and overheating
• Find processes consuming RAM or GPU memory

Prepare NVIDIA Jetson setup files by checking supported releases and downloading the BSP, sample root filesystem, sources, toolchain, and guides. Files are placed where the following setup skills expect to find them.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Preparing files for Jetson Auto Setup
• Gathering artifacts for a supported Jetson Linux release
• Staging downloads before image and source initialization

Write a prepared board support package image to a Jetson device in recovery mode. The skill checks device identity and image readiness, selects the flashing tool for the boot storage, and presents the resolved plan for approval.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Flashing a prepared image after customization
• Reflashing a Jetson with the same image
• Checking device identity and settings before flashing

Create a dated Markdown reference for an active Jetson target profile. It summarizes board support package and source directories, relevant configuration files, and document references, checking local paths so later sessions can consult a saved snapshot.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Documenting the layout after BSP extraction
• Refreshing the reference after source updates
• Checking configuration files for a target module

Prepare a Jetson BSP image, the board's system software, from local Jetson Linux and sample root filesystem archives. Apply GPU binaries for the selected target, verify existing images, and record image metadata in its profile.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Prepare a BSP image after selecting a Jetson target
• Extract matching Jetson Linux and root filesystem archives
• Check and reuse an existing Linux_for_Tegra directory

Prepare a workspace for customizing Jetson board support software. Sets up Git tracking for changes, arranges kernel and related source files, and locates or extracts the NVIDIA Crosstool-NG compiler tools needed for later builds.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Preparing a workspace after BSP image initialization
• Arranging source files from local archives for offline setup
• Using a customer Git fork as the BSP source

Create a Jetson target-platform YAML profile by selecting a reference product and flash configuration, with optional details for a custom carrier board. The workflow also sets the active profile used by subsequent setup and customization skills.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Creating the first Jetson target-platform profile
• Recording a custom carrier board and its reference platform
• Adding a profile for another Jetson product

Register existing Jetson and IGX reference documents in the active platform profile so downstream skills can find them by name. Suggest files based on filenames, confirm selections, and update references for guides, schematics, and pinmux spreadsheets.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Register reference documents after setting up a target profile
• Update or remove document references in an existing profile
• Fix missing document references reported by downstream skills

Measure language model response times and generation speed on Jetson using wrappers for vLLM, llama.cpp, or Ollama. Produces structured JSON results to assess models, power modes, runtime settings, and performance changes after tuning.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Measure response times for a deployed model
• Compare performance before and after tuning
• Evaluate models and power modes under controlled conditions

Provides instructions for serving language and vision-language models on Jetson with vLLM or SGLang. Selects a runtime for the device and JetPack version, supplies launch commands, and explains how to verify the OpenAI-compatible API.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Starting a language-model API on Jetson
• Connecting Open WebUI or an app to a model server
• Serving a vision-language model with vLLM

Reclaim application memory on Jetson deployments without displays or cameras. Guides changes to boot settings, kernel memory reservations, and the DMA transfer buffer, then checks compiled settings and memory usage on the running device.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Reclaiming memory on a Jetson without a display
• Removing camera memory reservations
• Shrinking and checking the DMA transfer buffer

Choose containers and Python package sources that match a Jetson device's GPU architecture, JetPack, and CUDA environment. Get compatibility hints for PyTorch, vLLM, and related stacks before installation, using NVIDIA catalogs and Jetson AI Lab indexes.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Choose an ML container for Jetson
• Find compatible PyTorch or vLLM package sources
• Investigate CUDA compatibility issues with ARM packages

Prepare a staged Jetson BSP image, the device software package, by copying customized files and built artifacts. Skip identical files and refresh startup files when kernels or modules change, without building or flashing the device.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Apply customized configuration files to a staged BSP image
• Copy built kernels and modules into the image
• Prepare an image for inspection before flashing

Configure speculative decoding on Jetson vLLM servers, letting a smaller model propose tokens for the main model to verify. Adjust memory allocation and benchmark before and after to assess generation speed under low concurrency.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Reducing slow token generation on Thor or AGX Orin
• Choosing EAGLE-3 or a small draft model
• Comparing generation speed before and after configuration

Plan, run, and verify Jetson video encoding, decoding, transcoding, and segment workflows using NVIDIA samples or PyNvVideoCodec. It checks file hashes, byte sizes, frame counts, and logs to confirm that each stage consumes the preceding stage’s output.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Verify an encode-to-decode workflow
• Check H.264-to-HEVC transcoding
• Decode and validate every output segment

Install, repair, or inspect NVIDIA Video Codec SDK and PyNvVideoCodec on Jetson. Verify readiness with official samples that encode and independently decode one frame, and report what passed, what failed, and the next action.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Install a video SDK on Jetson
• Inspect installed packages and the selected Python environment
• Verify one-frame encoding and decoding with official samples

Check and format Python code in Megatron-LM using autoformat.sh and its linting tools. Apply project rules for imports, type hints, documentation, naming, and line length when preparing changes for a pull request.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Check code style before opening a pull request
• Run automatic formatting fixes
• Sort imports after editing Python files

Launch distributed Megatron-LM training on a SLURM cluster with guidance on job scripts, environment variables, GPU settings, and containers. Monitor jobs and investigate failures across worker processes to identify the initial cause.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Write an sbatch script for distributed training
• Configure training across multiple GPU nodes
• Check CUDA settings for the parallelism mode

Split a pull request into smaller drafts based on CODEOWNERS, the file that assigns review ownership. Propose groups of changes that need fewer reviewer teams, keep tests with their code, and create the drafts after approval.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Split a large pull request into smaller reviews
• Reduce reviewer groups required for each pull request
• Plan dependencies between split pull requests

Add and run unit or functional tests for Megatron-LM. The guide explains test organization, YAML execution recipes, reference results for numerical checks, test filters, temporary disabling, and reproducing continuous integration failures locally.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Add unit tests or functional training tests
• Run selected tests and reproduce CI failures locally
• Download or update reference results for numerical checks

Configure how Dynamo routes requests to workers, then check model discovery and a chat response. Supports local and Kubernetes setups, including routing based on cached model data, with clear commands, results, and limitations.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Start a local Dynamo frontend
• Change routing settings on Kubernetes
• Check model discovery and a chat response

Check installed DOCA versions across package metadata, application files, runtime tools, and BlueField system information. Use the results to investigate build and runtime mismatches, check feature availability, and understand NGC container version tags.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Check whether installed DOCA versions agree
• Investigate build and runtime version mismatches
• Check feature support for a specific release

Assess whether a DOCA application needs low-level communication controls beyond its higher-level APIs. For those cases, guide queue configuration, libibverbs migration, and error diagnosis using installed headers, version checks, and device capability queries.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Decide whether higher-level DOCA APIs are sufficient
• Configure a queue attribute unavailable through the task API
• Port libibverbs code into DOCA Core

Helps developers offload remote memory operations from HPC, UCX, or MPI applications to a BlueField DPU, a processor for infrastructure tasks. Covers host-side DOCA UROM setup, supported-operation discovery, completion handling, and error investigation.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Offloading MPI communication work to a BlueField DPU
• Checking supported operations through plugin discovery
• Investigating errors from host-side UROM calls