High‑Performance Inference for PaddleOCR on Jetson ARM (aarch64)

[Zahrachab]

Hi Paddle team 👋, I’ve tested the High‑Performance Inference mode (enable_hpi True) on x86-64 Linux (observing ~25% speedup) and I'd like to do the same on Jetson ARM64 devices . Is native HPI support for ARM64 (e.g., Jetson Nano/Xavier/Orin) on your roadmap—perhaps as an ARM-compatible Docker image or extension to the CLI?

[Bobholamovic]

Currently, HPI does not support ARM64. Feel free to submit a feature request in issues, and we will consider adding support in future versions.