Install WebODM with GPU support

WebODM
,
1

Install WebODM with GPU support

Hello,
Has anyone already installed WebODM on Docker with GPU support using an NVIDIA RTX 3060 ZOTAR 12GB or similar on Windows 11 Pro and could you share a step-by-step guide even though it is not detailed about how you did it?
I installed it, but I’m suspicious that I’m making a mistake at some point, because I did two processings with the same dataset, under the same conditions and settings, except that I unchecked “no-gpu” in the first processing and checked it in the other; the one with “no-gpu” marked is almost twice as fast.
So, probably the way I installed it was wrong and when I disable the GPU the processing goes faster.

no-gpu
no-gpu1062×606 47.2 KB

no-gpu2

1 Like
2

Have you checked the logs to see if the GPU run was actually using the GPU?

1 Like
3

Also curious on this problem

1 Like
4

gpu

Yes. In the log file: “no_gpu”:false,

1 Like
5

But I think the problem is happening because I don’t know how to properly install the GPU.
I’m newbie at this!

1 Like
6

This test bellow return the error: Error: only 0 Devices available, 1 requested. Exiting.

root@photogrammetric:/home/ilton# cd WebODM
root@photogrammetric:/home/ilton/WebODM# docker run --gpus all nvcr.io/nvidia/k8s/cuda-sample:nbody nbody -gpu -benchmark
Unable to find image 'nvcr.io/nvidia/k8s/cuda-sample:nbody' locally
nbody: Pulling from nvidia/k8s/cuda-sample
f65423f1b49b: Download complete
207b64ab7ce6: Download complete
e9bff09d04df: Download complete
edc14edf1b04: Download complete
a424d45fd86f: Download complete
2b60900a3ea5: Download complete
1939e4248814: Download complete
1f37f461c076: Download complete
9026fb14bf88: Download complete
548afb82c856: Download complete
22c5ef60a68e: Download complete
Digest: sha256:59261e419d6d48a772aad5bb213f9f1588fcdb042b115ceb7166c89a51f03363
Status: Downloaded newer image for nvcr.io/nvidia/k8s/cuda-sample:nbody
Run "nbody -benchmark [-numbodies=<numBodies>]" to measure performance.
        -fullscreen       (run n-body simulation in fullscreen mode)
        -fp64             (use double precision floating point values for simulation)
        -hostmem          (stores simulation data in host memory)
        -benchmark        (run benchmark to measure performance)
        -numbodies=<N>    (number of bodies (>= 1) to run in simulation)
        -device=<d>       (where d=0,1,2.... for the CUDA device to use)
        -numdevices=<i>   (where i=(number of CUDA devices > 0) to use for simulation)
        -compare          (compares simulation results running once on the default GPU and once on the CPU)
        -cpu              (run n-body simulation on the CPU)
        -tipsy=<file.bin> (load a tipsy model file for simulation)

NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled.

Error: only 0 Devices available, 1 requested.  Exiting.
root@photogrammetric:/home/ilton/WebODM#
1 Like
7
  • Reasons for this error:
    1 No compatible GPU devices are present on your system.
    2 The CUDA toolkit is not properly installed or configured.
    3 The Docker image you’re using doesn’t support the specified GPU.

The only choice is the second case

1 Like
8 
root@photogrammetric:/home/ilton/WebODM# nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Thu_Nov_18_09:45:30_PST_2021
Cuda compilation tools, release 11.5, V11.5.119
Build cuda_11.5.r11.5/compiler.30672275_0
root@photogrammetric:/home/ilton/WebODM#
1 Like
9

Do you have a more recent CUDA release, GPU driver, firmware, and kernel to test under? CUDA v11.5 is quite a bit old.

10

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.