I have installed toolkit cuda 11.8 (arm64-Jetson)
jtop showes:
Release : 5.10.120-tegra
Jetpack : 5.1.2
L4T : 35.4.1
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Wed_Sep_21_10:43:33_PDT_2022
Cuda compilation tools, release 11.8, V11.8.89
Build cuda_11.8.r11.8/compiler.31833905_0
apt show nvidia-jetpack
Package: nvidia-jetpack
Version: 5.1.2-b104
Priority: standard
Section: metapackages
Maintainer: NVIDIA Corporation
Installed-Size: 199 kB
The LD_LIBRARAY_PATH is set to point to
/usr/local/cuda-11.8/compat
compiled and run natively.
sudo ./deviceQuery
cudaGetDeviceCount returned 35
→ CUDA driver version is insufficient for CUDA runtime version
Very poor , unorganized , all over the place , contradicting , unchecked … documentation
Worst I have ever seen, even worse than Xilinx.
Looks like Nvidia staff has no interest to get to the butom of issues.
Tried removal , reinstall …
Hello,
Thanks for visiting the NVIDIA Developer Forums.
To ensure better visibility and support, I’ve moved your post to the Jetson category where it’s more appropriate
Cheers,
Tom
Hi,
JetPack 5 is released years ago.
Do you have any dependencies on it?
If not, for AGX Orin, it’s more recommended to use JetPack 6.2.2 with CUDA 12.6 instead.
Thanks.
Hi,
Can you clarify the statement “Do you have any dependencies on it?”
it = cuda 11.8 ? it = orin ?
The 11.8 has been praised by Nvidia as a great quality release.
I intend to use cuda and dpdk on orin, if this is a dependency.
I have the dpdk working with a Nvidia NIC.
Hi,
Do you have any dependencies on JetPack 5?
If not, please upgrade to the latest software release.
Thanks.
I have tried to flash the Jetson Orin AGX to the latest jetpack. My system is behind a proxy.
The sdk manager fails to finish logging in. I have entered the proxy settings in SDK Manager, installed, reinstalled and numerous other tries suggested. I also tried to do the flashing from command line and sheel scripts I got from github. The results with SDK Manager is either session timeout , or token could not be validated. I am able to login to the nvidia development and also see the the SDK Manager logs in as well.
Is Nvidia hoping that these basic primitive issues resolve themselves on their own?
It seems that no one cares in Nvidia. Is Nvidia management aware of these socking low qulity
software?