Cannot Complete SDK Install

Hello,

I am installing the newest version L4T and JetPack onto a TX1. During the installation process I am getting this error:

22:59:57 INFO: cuDNN on Target: dpkg: error processing archive /opt/nvidia/deb_repos/libcudnn8_8.0.0.180-1+cuda10.2_arm64.deb (–install):

22:59:57 INFO: cuDNN on Target: package architecture (arm64) does not match system (amd64)

22:59:58 INFO: cuDNN on Target: Errors were encountered while processing:

22:59:58 INFO: cuDNN on Target: /opt/nvidia/deb_repos/libcudnn8_8.0.0.180-1+cuda10.2_arm64.deb

22:59:58 INFO: cuDNN on Target: Unhandled error when running sudo dpkg -i /opt/nvidia/deb_repos/libcudnn8_8.0.0.180-1+cuda10.2_arm64.deb

22:59:58 INFO: cuDNN on Target: [ Package Install Finished with Error ]

22:59:58 INFO: cuDNN on Target: [ NV_L4T_CUDNN_TARGET_POST_INSTALL_COMP Install took 3s ]

22:59:58 ERROR: cuDNN on Target: [error]: dpkg install on target failed; [exec_command]: safe_apt_install “sudo dpkg -i /opt/nvidia/deb_repos/libcudnn8_8.0.0.180-1+cuda10.2_arm64.deb” 5; sleep 5; ; [deb_path]: /home/dylan/Downloads/nvidia/sdkm_downloads/libcudnn8_8.0.0.180-1+cuda10.2_arm64.deb

22:59:58 ERROR: cuDNN on Target: command terminated with error

22:59:58 SUMMARY: cuDNN on Target: Failed to install debian package [target]

22:59:58 SUMMARY: NVIDIA Container Runtime with Docker integration (Beta): Depends on failed component

22:59:58 SUMMARY: Multimedia API: Depends on failed component

22:59:58 SUMMARY: TensorRT on Target: Depends on failed component

22:59:58 SUMMARY: NVIDIA Container Runtime with Docker integration (Beta): Depends on failed component

22:59:58 SUMMARY: Multimedia API: Depends on failed component

22:59:58 SUMMARY: VisionWorks on Target: Install completed successfully.

What is going on here?

Hi djtussey,

Are you using Ubuntu 16.04/18.04 host machine to install JetPack?
Please using the same version to flash image and install sdk components.

22:59:57 INFO: cuDNN on Target: package architecture (arm64) does not match system (amd64)

I restarted the process from the SDK manager and was able to successfully get my Jetson updated. Thank you!

1 Like