I am having issues getting webodm to recognize my nvidia rtx 2060. Any help would be appreciated nvidia-smi not found in PATH, using CPU. Please help !
2 Likes
What is the output of :
docker run --rm --gpus all nvidia/cuda:10.0-base nvidia-smi
1 Like
±----------------------------------------------------------------------------+
| NVIDIA-SMI 510.47.03 Driver Version: 511.65 CUDA Version: 11.6 |
|-------------------------------±---------------------±---------------------+
| 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 NVIDIA GeForce … On | 00000000:01:00.0 On | N/A |
| 29% 33C P8 19W / 125W | 599MiB / 6144MiB | 1% Default |
| | | N/A |
±------------------------------±---------------------±---------------------+
±----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
±----------------------------------------------------------------------------+
1 Like
Not sure what I am doing wrong.
1 Like
Have you checked the guidance here:
I did but how is that implemented into a odm docker?
1 Like
Docker for Windows should be using the WSL2 backend by default, so you will need to meet all the various requirements for WSL2 CUDA GPU Passthrough for ODM on Docker on Windows to have access to the GPU.
By installing the nvidia docker ?
I used windows git and pulled the repository to and then used git bash to push the docker image out.
1 Like
Try this
1 Like
Windowed mode
Simulation data stored in video memory
Single precision floating point simulation
1 Devices used for simulation
GPU Device 0: “Turing” with compute capability 7.5
Compute 7.5 CUDA device: [NVIDIA GeForce RTX 2060]
30720 bodies, total time for 10 iterations: 48.560 ms
= 194.342 billion interactions per second
= 3886.839 single-precision GFLOP/s at 20 flops per interaction
1 Like
when I run
“./webodm.sh --gpu”
Warning: GPU support is not available for Windows
Usage: ./webodm.sh
1 Like
It will always say that but when processing it will be used at start. You can check task manager and click on GPU at start for extracting features.
I hope you changed files accordingly to that thread.
1 Like
I looked in the file but it seemed to be what was already there?
1 Like
Example left mine on the right
1 Like
Steps 9 and 10
Just copy the whole thing.
Difference is at the end. And in webodm.sh there’s just one change to true.
2 Likes
Oohh, totally looked over that one. Sorry.
1 Like
Good sign? It decided to randomly download more?
1 Like
Yup. Should be fine. Download is around 2gb…
Then you should always start with --gpu. You will see if it works from the very start of process pipeline.
2 Likes
Gonna test a small dataset really quick and will let you know.
1 Like
It’s working, Thank you so much!!!
2 Likes