GPU Pipeline: disabled, Not connected to PVD
Opened this issue · 5 comments
I tried the examples “python joint_monkey.py” and got error. It told me that GPU Pipeline is disabled. It’s on Ubuntu 18.04.1 and RTX3090.
How can i enable my GPU pipeline?
Thankful for your help!
Error message as followings:
Importing module 'gym_37' (/home/robo/isaacgym/python/isaacgym/_bindings/linux-x86_64/gym_37.so)
Setting GYM_USD_PLUG_INFO_PATH to /home/robo/isaacgym/python/isaacgym/_bindings/linux-x86_64/usd/plugInfo.json
WARNING: Forcing CPU pipeline.
Not connected to PVD
+++ Using GPU PhysX
Physics Engine: PhysX
Physics Device: cuda:0
GPU Pipeline: disabled
Segmentation fault (core dumped)
The result for “nvidia-smi” is as followings:
Fri Feb 24 10:59:23 2023
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.89.02 Driver Version: 525.89.02 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| 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 ... Off | 00000000:65:00.0 On | N/A |
| 32% 35C P8 29W / 350W | 258MiB / 24576MiB | 12% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 1415 G /usr/lib/xorg/Xorg 20MiB |
| 0 N/A N/A 2162 G /usr/lib/xorg/Xorg 235MiB |
+-----------------------------------------------------------------------------+
And the cuda 11.6 is also available in pytorch:
Python 3.7.12 | packaged by conda-forge | (default, Oct 26 2021, 06:08:53)
[GCC 9.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
import torch
print(torch.version.cuda)
11.6
I have no idea how to fix this problem.
Hello, I have also encountered the same problem. Have you resolved this issue?
Facing the same issue... any solution yet?
have you solved this problem?
Has anyone been able to solve this?
I was able to solve this by installing nvcc
sudo apt install nvidia-cuda-toolkit