Skip to main content

NVIDIA

NVIDIA GPUs with NVENC provide hardware-accelerated video encoding that's 5-10x faster than software encoding. If you have a GTX 1050 or newer, you've got NVENC.

What you get​

  • NVENC encoding: h264_nvenc, hevc_nvenc. Offloads encoding to dedicated hardware on the GPU.
  • NVDEC decoding: hardware-accelerated decode, keeps the full pipeline on GPU.
  • CUDA scaling: scale_cuda resizes frames on the GPU instead of pulling them back to CPU.
  • OpenCV CUDA face detection: face-aware cropping runs on the GPU. Faster than CPU, though not as fast as Apple's Neural Engine.

Requirements​

  • NVIDIA GPU (GTX 1050+ / any RTX)
  • CUDA drivers installed
  • FFmpeg built with NVENC support (most distro packages include this)

Check if everything's working:

immich-memories hardware

If NVENC is available, you'll see it listed with the specific encoders found.

Configuration​

hardware:
  enabled: true
  backend: "nvidia"   # or "auto" (auto will find it)

You usually don't need to set backend: "nvidia" explicitly. auto detects NVIDIA GPUs fine. The only reason to force it is if you have multiple acceleration options and want to ensure NVIDIA is used.

Encoding quality​

NVENC quality is slightly below software libx264 at the same bitrate, but for memory videos the difference is invisible. The speed gain (5-10x) is worth it. If you're encoding a 2-minute compilation, NVENC finishes in seconds instead of minutes.