Please provide the following info (tick the boxes after creating this topic): Software Version
[*] DRIVE OS 6.0.10.0
DRIVE OS 6.0.8.1
DRIVE OS 6.0.6
DRIVE OS 6.0.5
DRIVE OS 6.0.4 (rev. 1)
DRIVE OS 6.0.4 SDK
other
Target Operating System
[*] Linux
QNX
other
Hardware Platform
DRIVE AGX Orin Developer Kit (940-63710-0010-300)
DRIVE AGX Orin Developer Kit (940-63710-0010-200)
[*] DRIVE AGX Orin Developer Kit (940-63710-0010-100)
DRIVE AGX Orin Developer Kit (940-63710-0010-D00)
DRIVE AGX Orin Developer Kit (940-63710-0010-C00)
DRIVE AGX Orin Developer Kit (not sure its number)
other
SDK Manager Version
2.1.0
other
Host Machine Version
native Ubuntu Linux 20.04 Host installed with SDK Manager
[*] native Ubuntu Linux 20.04 Host installed with DRIVE OS Docker Containers
native Ubuntu Linux 18.04 Host installed with DRIVE OS Docker Containers
other
Issue Description
I am trying to validate the GPU is enabled or not on Nvidia Drive AGX Orin Platform for which I used below commands :
a) In Python environment - torch.cuda.is_available() which shows result as False
b) and check output of nvidia-smi but this bin is not present in the system.
Hence, I followed the instructions in the link CUDA Toolkit 12.6 Update 2 Downloads | NVIDIA Developer
and after performing sudo apt-get install -y cuda-drivers the system shows unmet dependencies and unable to remove any packages nor upgrade any packages. Also dpkg --configure -a fails.
What is the way to recover the system instead of flashing?
Attaching the complete error logs while updating/upgrading or removing any packages. cuda_installation_error_logs.txt (60.0 KB)
If you want to test if the GPU is detected and no issue with GPU on Orin devkit, you can use deviceQuery .
nvidia@tegra-ubuntu:~$ cd /usr/local/cuda-11.4/samples/1_Utilities/deviceQuery
nvidia@tegra-ubuntu:/usr/local/cuda-11.4/samples/1_Utilities/deviceQuery$ ls
Makefile NsightEclipse.xml deviceQuery deviceQuery.cpp deviceQuery.o readme.txt
nvidia@tegra-ubuntu:/usr/local/cuda-11.4/samples/1_Utilities/deviceQuery$ ./deviceQuery
./deviceQuery Starting...
CUDA Device Query (Runtime API) version (CUDART static linking)
Detected 1 CUDA Capable device(s)
Device 0: "Orin"
CUDA Driver Version / Runtime Version 12.1 / 11.4
CUDA Capability Major/Minor version number: 8.7
Total amount of global memory: 28954 MBytes (30360248320 bytes)
(016) Multiprocessors, (128) CUDA Cores/MP: 2048 CUDA Cores
GPU Max Clock rate: 1275 MHz (1.27 GHz)
Memory Clock rate: 1275 Mhz
Memory Bus Width: 256-bit
L2 Cache Size: 4194304 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 167936 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: No
Integrated GPU sharing Host Memory: Yes
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 0 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.1, CUDA Runtime Version = 11.4, NumDevs = 1
Result = PASS
The DRIVE OS 6.0.10 comes with CUDA 11.4 already and you are not expected to install CUDA toolkit or drivers on target.
Just an update I have the package manager which was broken has been recovered by below commands :
a) sudo dpkg --force-all -P nvidia-compute-utils-560 nvidia-container-toolkit nvidia-driver-560 libnvidia-compute-560 libnvidia-container-tools libnvidia-gl-560 libnvidia-extra-560 libnvidia-decode-560 libnvidia-encode-560 nvidia-utils-560 libnvidia-cfg1-560 libnvidia-fbc1-560 libnvidia-cfg1-560 xserver-xorg-video-nvidia-560
b) sudo apt autoremove && sudo apt autoclean && sudo apt clean && sudo apt-get update && sudo apt upgrade
c) sudo dpkg --force-all -P cuda-drivers cuda-drivers-560 nvidia-driver-560 nvidia-driver-560-open nvidia-driver-560-server nvidia-driver-560-server-open
d) sudo dpkg --configure -a
e) sudo apt --fix-broken install && sudo apt autoremove
Thanks for sharing the inputs. Able to get the same output as shared. But as I do not find nvidia-smi utility on the system to check what can be done further. To get this utility only I tried to install cuda-toolkit.
Also are there any steps to force an application to execute on GPU? I have an Yolo based application which reads the camera images in PNG format and do the annotation on top of that and convert again into PNG image. I am executing this Yolo application in conda environment. How to ensure to run this application on GPU?
Note that CUDA 11.4 installed on target is officially supported on DRIVE. TensorRT and DW are expected to use CUDA 11.4.
Are you using TensorRT or DW application or something else? Please provide details of the application in a new topic to avoid cluttering of issues in a single topic.
