Guidance on Building FFMPEG with NVIDIA GPU Support for H.265 Decoding

Robotics & Edge Computing Jetson & Embedded Systems Jetson AGX Orin
1

Description

Thermal spike during ffmpeg image decode on Jetson

Screenshot from 2025-04-22 15-40-31
Screenshot from 2025-04-22 15-40-31906×672 108 KB

Environment

TensorRT Version: 10.3.0.30
GPU Type: Tegra
Nvidia Driver Version: 540.4.0
CUDA Version: 12.6.68
CUDNN Version: 9.3.0.75
Operating System + Version: Ubuntu 22.04
Python Version (if applicable): 3.10.12
Baremetal or Container (if container which image + tag): ultralytics/ultralytics:latest-jetson-jetpack6

Usecase: Extracting images from 4 minute GoPro downloaded video of resolution (5K) using ffmpeg on Jetson AGX Orin.

Problem: ffmpeg library takes good chunk of time on CPU compute and increase jetson thermal over the period of time during the extraction process.

Would like to request detailed guidance on building optimized FFMPEG for jetson with full NVIDIA GPU acceleration support particularly for H.265 (HEVC) encoding and decoding using NVENC/NVDEC (or CUVID) to speedup extraction workload on GPU.

2

Hi @arumugamdeepakkumar ,
Please reach out to Jetson Forum for better assistance.
Thank you.

4

Hi,
Please refer to steps in FAQ and give it a try:
Jetson AGX Orin FAQ

And please note the encoding/decoding is on individual NVENC and NVDEC engines instead of GPU.