Cross Operating System platform compatibility with the same GPU platform

A quick question:
I have compiled a TensorRT inference engine on Linux x86-64 Ubuntu 18.04.2 LTS and GPU 3070.

  1. Is it ok to reuse it on Windows-7 x86-64 and the same GPU 3070, plus the same (or as similar as possible) TensorRT and CUDA environment?
  2. If not, any issue or problem?
  3. And what can I do about it?



The generated engine files are not portable across platforms or TensorRT versions. TRT engine files are specific to the exact GPU model they were built on (in addition to the platforms and the TensorRT version, OS, etc.) and must be re-built on the platform you’re planning to deploy.

Thank you.

1 Like