Hi Yves,, 

I don't see a big problem here. The -arch option is suspicious, since it 
should be *compute_XX*, like *compute_80*, to specify a compute capability. 
But why it appears not detected or how it interacts with using Jupyter, I 
do not know, since I don't use it. 

If it's about that your device is not detected while using Jupyter, then I 
don't know if I can reproduce that. I can have a try, but unable to 
guarantee. I would suggest that you try if you can run any other CUDA-based 
packages using Jupyter. BTW I don't know why trimesh could have this 
problem and it's weird.

Ruochun

On Friday, March 22, 2024 at 10:04:22 PM UTC+8 [email protected] wrote:

> Hi Ruochun,
>
> It is quite difficult to install the tool given the cuda version to use. I 
> made it after trying several combinations for a long time.
> The following is a script I made to install on Ubuntu 22.04 and my type of 
> graphic cards (on a fresh instance):
>
> #!/bin/bash
>
> # Remove old versions of CUDA and NVIDIA tools
> sudo apt-get --purge remove -y "cublas*" "cuda*" "nvidia*"
> sudo rm -rf /usr/local/cuda*
> sudo apt-get autoremove -y && sudo apt-get autoclean -y
>
> # Install necessary utilities and libraries
> sudo apt-get update
> sudo apt-get install -y g++ freeglut3-dev build-essential libx11-dev 
> libxmu-dev libxi-dev libglu1-mesa libglu1-mesa-dev
>
> # Install new CUDA, version 12.0 (amd, Ubuntu 22.04)
> wget 
> https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin
> sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600
> wget 
> https://developer.download.nvidia.com/compute/cuda/12.0.0/local_installers/cuda-repo-ubuntu2204-12-0-local_12.0.0-525.60.13-1_amd64.deb
> sudo dpkg -i cuda-repo-ubuntu2204-12-0-local_12.0.0-525.60.13-1_amd64.deb
> sudo cp /var/cuda-repo-ubuntu2204-12-0-local/cuda-*-keyring.gpg 
> /usr/share/keyrings/
> sudo apt-get update
> sudo apt-get install -y cuda-12.0
>
> # Update environment variables
> echo 'export PATH=/usr/local/cuda-12.0/bin:$PATH' >> ~/.bashrc
> echo 'export LD_LIBRARY_PATH=/usr/local/cuda-12.0/lib64:$LD_LIBRARY_PATH' 
> >> ~/.bashrc
>
> # Install Miniconda
> wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
> bash Miniconda3-latest-Linux-x86_64.sh -b -p $HOME/miniconda
>
> # Initialize Conda for the shell
> $HOME/miniconda/bin/conda init
> echo 'export PATH="$HOME/miniconda/bin:$PATH"' >> ~/.bashrc
>
> # Activate conda environment
> source $HOME/.bashrc
> source $HOME/miniconda/bin/activate
>
> # Install conda environment and tools
> conda create -y -n pyDEME python=3.11
> conda activate pyDEME
> conda install -y cmake
> pip3 install DEME
>
> # OPTIONAL: few tools I use
> conda config --add channels conda-forge
> conda install -y seaborn matplotlib numpy scipy trimesh paraview shapely 
> gmsh PyArrow
> pip install gmsh
>
> echo "Installation completed. Please reboot your system to apply all 
> changes and for the CUDA installation to be fully operational."
>
> I am not sure what exactly I am missing, to be honest.
>
> Best regards,
> Yves
> On Friday, March 22, 2024 at 4:42:02 AM UTC-4 Ruochun Zhang wrote:
>
>> Hi Yves,
>>
>> If you say this also happens to other packages like trimesh which is 
>> fully CPU-based, then I suspect it's the C++ compiler version. C++11 is way 
>> too old for anything. You should try updating the compiler that you are 
>> using for the installation of all the packages so it supports newer 
>> standards, at least C++17. That is my best guess at this moment.
>>
>> Thank you,
>> Ruochun
>>
>> On Wednesday, March 20, 2024 at 5:04:16 AM UTC+8 [email protected] 
>> wrote:
>>
>>> Hello,
>>>
>>> I would like to know how to run DEM-Engine within a Jupyter notebook.
>>> Indeed, I obtain the following error:
>>>
>>> Compiler options: -diag-suppress=550 -diag-suppress=177 -arch=compute_ 
>>> -std=c++11 
>>> Traceback (most recent call last):
>>>   File "/home/cloud/PBR_DEME/test.py", line 46, in <module>
>>>     S.Instance.Initialize()
>>> RuntimeError: NVRTC error: NVRTC_ERROR_INVALID_OPTION
>>>
>>> I also face this- same issue when using the trimesh python module 
>>> outside of Jupyter. However, I would like to know what triggers that error 
>>> so that I will hopefully solve both issues.
>>>
>>> Thanks!
>>>
>>

-- 
You received this message because you are subscribed to the Google Groups 
"ProjectChrono" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/projectchrono/26302c5d-e44b-4bd2-a80a-3a02afcbd975n%40googlegroups.com.

Reply via email to