Error with Quadro M1000M: CUDA Kernel Image Not Available for Execution

[Xu]

I’m trying to run my old computer with a Quadro M1000M to complete a parallel simulation, but the process is aborted right after the simulation starts, and I get the following error message:

terminate called after throwing an instance of 'thrust::THRUST_200500_500_NS::system::system_error'
  what():  parallel_for failed: cudaErrorNoKernelImageForDevice: no kernel image is available for execution on the device
Aborted

Does this indicate that my graphics card is not supported for parallel simulations?

[Adrian]

No, your GPU is supported by CUDA with CUDA_ARCH set to 50 in our config.mk. As the error originates not from OpenLB but from the NVIDIA Thrust library we use in some parts internally the problem could be with your CUDA installation. Can you show a backtrace and show your config? Did you successfully run other CUDA applications on this machine?

[Xu]

It actually makes me kinda confused since both CUDA and CUDA toolkit should have been installed successfully, and I had run same simulation with same method couple weeks ago on the same computer before i move Ubuntu to another HD Drive, maybe I should try re-install OpenLB?

xu@XUS-P50-OF:~/openlb/olb-1.7r0/examples/multiComponent/phaseSeparation3d$ nvidia-smi
Fri Jan 10 09:20:27 2025       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.01             Driver Version: 553.50       CUDA Version: 12.4     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  Quadro M1000M                  On  | 00000000:01:00.0  On |                  N/A |
| N/A    0C    P5              N/A / 200W |    871MiB /  2048MiB |      4%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
|    0   N/A  N/A       109      G   /Xwayland                                 N/A      |
+---------------------------------------------------------------------------------------+

nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Tue_Oct_29_23:50:19_PDT_2024
Cuda compilation tools, release 12.6, V12.6.85
Build cuda_12.6.r12.6/compiler.35059454_0

[Xu]

Sorry, is it because my CUDA version and tool version not matched well?

[Adrian]

Yes, it seems so.

[Xu]

thanks a lot!

[Author]

Posts