Syllo
nvtop --No GPU to monitor.
Hi after reviewing most of the suggestions from MS and NV, and surfing the web, I can not get nvtop to see my GPU. Ollama saw the GPU and used but nvtop failed to see it?
Here is some information about the setup of my machine:
wsl --version Versión de WSL: 2.3.24.0 Versión de kernel: 5.15.153.1-2 Versión de WSLg: 1.0.65 Versión de MSRDC: 1.2.5620 Versión de Direct3D: 1.611.1-81528511 Versión DXCore: 10.0.26100.1-240331-1435.ge-release Versión de Windows: 10.0.22631.4169
$ 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
$ nvidia-smi Fri Oct 4 13:57:04 2024 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 565.51.01 Driver Version: 565.90 CUDA Version: 12.7 | |-----------------------------------------+------------------------+----------------------+ | 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 RTX 4090 On | 00000000:05:00.0 On | Off | | 0% 49C P8 10W / 450W | 1536MiB / 24564MiB | 5% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+----------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=====================================================================================| | 0 N/A N/A 26 G /Xwayland N/A | +-----------------------------------------------------------------------------------------+
$ curl -fsSL https://ollama.com/install.sh | sh
Installing ollama to /usr/local Downloading Linux amd64 bundle ######################################################################## 100.0%##O=# # ######################################################################## 100.0% Adding ollama user to render group... Adding ollama user to video group... Adding current user to ollama group... Creating ollama systemd service... Enabling and starting ollama service... Nvidia GPU detected. The Ollama API is now available at 127.0.0.1:11434. Install complete. Run "ollama" from the command line. mywsl@Galileo:~$ ollama run llama3.2:3b Good morning Good morning! How can I assist you today?
$ nvtop -v -> nvtop version 1.2.2 $ nvtop -> No GPU to monitor.
Do you have any suggestions about how to fix the problem? Have I missed something? Many thanks for considering my request.
Your Kernel might need to be upgraded. My system is Ubuntu, and this works for me :
$ sudo apt remove nvtop
$ sudo reboot to update to a new kernel
then reboot the system by power off so that the new kernel can be loaded
$ sudo apt install nvtop
Thanks for your answer. I did what you suggested but still the same error.
~$ nvtop No GPU to monitor. ~$ nvtop -v nvtop version 3.0.0
Thanks for your answer. I did what you suggested but still the same error.
~$ nvtop No GPU to monitor. ~$ nvtop -v nvtop version 3.0.0
i am curious if you have windows 10 or 11?
Windows 11 Pro.
i tried both, i got the same error when i used win10 but it worked fine when i used windows 11 pro.
Same issue. Windows 11 Pro, 2 Nvidia GPUs. I am running on WSL2 Ubuntu 22.04.3 LTS. "watch nvidia-smi" shows GPUs correctly. But nvtop responds with "No GPU to monitor." NVML appears to see and report my GPUs correctly. I've run through all of the reinstalling checks, but can't figure it out. Is it Ubuntu 22.04.3? WSL2?
~$ nvtop -v --> nvtop version 3.1.0
Possibly --> no WSL2 kernel support for /dev/nvidia*
nvtop-git latest still no luck on Arch Linux WSL
[100%] Built target nvtop
❯ sudo make install
[100%] Built target nvtop
Install the project...
-- Install configuration: "Release"
-- Installing: /usr/local/bin/nvtop
-- Set non-toolchain portion of runtime path of "/usr/local/bin/nvtop" to "/usr/local/lib"
-- Installing: /usr/local/share/man/man1/nvtop.1
-- Installing: /usr/local/share/icons/nvtop.svg
-- Installing: /usr/local/share/applications/nvtop.desktop
-- Installing: /usr/local/share/metainfo/nvtop.metainfo.xml
❯ nvtop
No GPU to monitor.
I'm having the same issue. my ollama docker container running from my Ubuntu 24.04.1 LTS WSL distro on a windows 11 PC is seeing the GPU just fine and utilizing it. nvidia-smi on ubuntu also sees the GPU. it's just nvtop (v3.0.2) which is showing "no GPU to monitor"
Same. nvtop 3.0.2 Ubuntu 24.04.1 LTS WSL2 on Windows 11 Pro
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 565.77.01 Driver Version: 566.36 CUDA Version: 12.7 |
|-----------------------------------------+------------------------+----------------------+
| 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 RTX 2080 Ti On | 00000000:01:00.0 On | N/A |
| 0% 34C P8 28W / 300W | 1317MiB / 11264MiB | 15% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| No running processes found |
+-----------------------------------------------------------------------------------------+
It looks like this just started happening on Arch Linux as well. Running nvidia-open 570.86.16-3.1.
It worked on the 6th..
But I'm noticing that even nvidia-smi is giving me:
Failed to initialize NVML: Driver/library version mismatch
NVML library version: 570.86
And that worked on Jan 28.
My GPU is def working.
/opt/cuda/extras/demo_suite/bandwidthTest
[CUDA Bandwidth Test] - Starting...
Running on...
Device 0: NVIDIA RTX A4000
Quick Mode
Host to Device Bandwidth, 1 Device(s)
PINNED Memory Transfers
Transfer Size (Bytes) Bandwidth(MB/s)
33554432 2904.0
Device to Host Bandwidth, 1 Device(s)
PINNED Memory Transfers
Transfer Size (Bytes) Bandwidth(MB/s)
33554432 2729.6
Device to Device Bandwidth, 1 Device(s)
PINNED Memory Transfers
Transfer Size (Bytes) Bandwidth(MB/s)
33554432 397060.8
Result = PASS
NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled.
Maybe nvidia broke something
Hi after reviewing most of the suggestions from MS and NV, and surfing the web, I can not get nvtop to see my GPU. Ollama saw the GPU and used but nvtop failed to see it?
Here is some information about the setup of my machine:
wsl --version Versión de WSL: 2.3.24.0 Versión de kernel: 5.15.153.1-2 Versión de WSLg: 1.0.65 Versión de MSRDC: 1.2.5620 Versión de Direct3D: 1.611.1-81528511 Versión DXCore: 10.0.26100.1-240331-1435.ge-release Versión de Windows: 10.0.22631.4169
$ 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
$ nvidia-smi Fri Oct 4 13:57:04 2024 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 565.51.01 Driver Version: 565.90 CUDA Version: 12.7 | |-----------------------------------------+------------------------+----------------------+ | 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 RTX 4090 On | 00000000:05:00.0 On | Off | | 0% 49C P8 10W / 450W | 1536MiB / 24564MiB | 5% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+----------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=====================================================================================| | 0 N/A N/A 26 G /Xwayland N/A | +-----------------------------------------------------------------------------------------+
$ curl -fsSL https://ollama.com/install.sh | sh
Installing ollama to /usr/local Downloading Linux amd64 bundle ######################################################################## 100.0%##O=# # ######################################################################## 100.0% Adding ollama user to render group... Adding ollama user to video group... Adding current user to ollama group... Creating ollama systemd service... Enabling and starting ollama service... Nvidia GPU detected. The Ollama API is now available at 127.0.0.1:11434. Install complete. Run "ollama" from the command line. mywsl@Galileo:~$ ollama run llama3.2:3b Good morning Good morning! How can I assist you today?
$ nvtop -v -> nvtop version 1.2.2 $ nvtop -> No GPU to monitor.
Do you have any suggestions about how to fix the problem? Have I missed something? Many thanks for considering my request.
I think there was no profiling was enabled
I'm having the same issue. my ollama docker container running from my Ubuntu 24.04.1 LTS WSL distro on a windows 11 PC is seeing the GPU just fine and utilizing it. nvidia-smi on ubuntu also sees the GPU. it's just nvtop (v3.0.2) which is showing "no GPU to monitor"
Mostly the profling has not update or else use the latest nvtop version
I also have the same issue (nvtop installed via brew install nvtop)
I'm working on:
lsb_release -a
lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 20.04.6 LTS
Release: 20.04
Codename: focal
nvtop info:
nvtop
# NO GPU to monitor.
nvtop --verison
# 3.2.0
however, nvidia-smi works fine:
nvidia-smi
Sun Mar 30 10:51:42 2025
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.01 Driver Version: 535.183.01 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| 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 RTX 4070 Ti Off | 00000000:01:00.0 On | N/A |
| 0% 35C P8 8W / 285W | 372MiB / 12282MiB | 0% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
+---------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=======================================================================================|
| 0 N/A N/A 1581 G /usr/lib/xorg/Xorg 35MiB |
| 0 N/A N/A 3561 G /usr/lib/xorg/Xorg 190MiB |
| 0 N/A N/A 3877 G /usr/bin/gnome-shell 29MiB |
| 0 N/A N/A 14632 G ...ures=SpareRendererForSitePerProcess 18MiB |
| 0 N/A N/A 80357 G ...seed-version=20250207-050113.109000 22MiB |
| 0 N/A N/A 2172711 G /usr/bin/totem 17MiB |
| 0 N/A N/A 4157827 G ...yOnDemand --variations-seed-version 25MiB |
+---------------------------------------------------------------------------------------+
I was able to fix mine. When 6.13 came out on arch, nvidia drivers were busted, I had to get some early-leaked drivers to fix it (wget https://github.com/flathub/org.freedesktop.Platform.GL.nvidia/releases/download/cuda/NVIDIA-Linux-x86_64-570.86.10.run I think), and I had never installed nvidia drivers from one of their .run files before.. and I actually thought it was pretty nice.. and the dkms install part seemed to be some extra separate step which put me out of sync with later drivers that I installed via pacman.
I was able to get things back on track just by downloading the latest .run installer, and using that again.
I’m experiencing the same issue on Arch Linux, but nvitop works and might be a good alternative.
https://github.com/XuehaiPan/nvitop
Make sure you don't have anything from a system package manager, like libnvidia-compute- packages in Ubuntu. You may have accidentally install them with nvidia-smi coming from apt. You should only use a bundle coming from the Windows drivers in /usr/lib/wsl/lib/ ref: https://docs.nvidia.com/cuda/wsl-user-guide/index.html
The only problem I have is wrong memory usage numbers.
Ubuntu from msstore: Ubuntu 24.04.2 LTS
I've built nvtop from source:
cmake .. -DAMDGPU_SUPPORT=OFF -DINTEL_SUPPORT=OFF -DMSM_SUPPORT=OFF -DAPPLE_SUPPORT=OFF -DPANFROST_SUPPORT=OFF -DPANTHOR_SUPPORT=OFF -DASCEND_SUPPORT=OFF -DV3D_SUPPORT=OFF -DTPU_SUPPORT=OFF
$ nvtop --version
nvtop version 3.2.0
Hi everyone,
I think I have found a possible cause for the "No GPU to monitor" issue in WSL. nvtop works fine for me on WSL until I install nvidia-cuda-toolkit. After installing nvidia-cuda-toolkit, running nvtop gives the "No GPU to monitor" error.
I tested this multiple times:
Without nvidia-cuda-toolkit: nvtop works and detects my GPU. After installing nvidia-cuda-toolkit: nvtop shows "No GPU to monitor". It seems that installing nvidia-cuda-toolkit somehow causes this problem. Maybe it's related to the drivers or libraries that the toolkit package installs or overwrites.
I hope this information helps to troubleshoot the issue!
After installing nvidia-cuda-toolkit, running nvtop gives the "No GPU to monitor" error.
From https://docs.nvidia.com/cuda/wsl-user-guide/index.html#cuda-support-for-wsl-2
Once a Windows NVIDIA GPU driver is installed on the system, CUDA becomes available within WSL 2. The CUDA driver installed on Windows host will be stubbed inside the WSL 2 as libcuda.so, therefore users must not install any NVIDIA GPU Linux driver within WSL 2. One has to be very careful here as the default CUDA Toolkit comes packaged with a driver, and it is easy to overwrite the WSL 2 NVIDIA driver with the default installation. We recommend developers to use a separate CUDA Toolkit for WSL 2 (Ubuntu) available from the CUDA Toolkit Downloads page to avoid this overwriting. This WSL-Ubuntu CUDA toolkit installer will not overwrite the NVIDIA driver that was already mapped into the WSL 2 environment. To learn how to compile CUDA applications, please read the CUDA documentation for Linux.
Hi everyone,
I think I have found a possible cause for the "No GPU to monitor" issue in WSL. nvtop works fine for me on WSL until I install nvidia-cuda-toolkit. After installing nvidia-cuda-toolkit, running nvtop gives the "No GPU to monitor" error.
I tested this multiple times:
Without nvidia-cuda-toolkit: nvtop works and detects my GPU. After installing nvidia-cuda-toolkit: nvtop shows "No GPU to monitor". It seems that installing nvidia-cuda-toolkit somehow causes this problem. Maybe it's related to the drivers or libraries that the toolkit package installs or overwrites.
I hope this information helps to troubleshoot the issue!
thanks, I can confirm work, I only install toolkit sudo ./cuda_12.9.1_575.57.08_linux.run --silent --toolkit --toolkitpath=/opt/cuda-12.9.1 --override
Same original problem here .... "No GPU to monitor"
- Ubuntu Server 24.04 LTS
- NVIDIA GeForce RTX 3060.
- NVIDIA driver 570.158.01
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 570.158.01 Driver Version: 570.158.01 CUDA Version: 12.8 |
|-----------------------------------------+------------------------+----------------------+
| 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 RTX 3060 Off | 00000000:2D:00.0 Off | N/A |
| 0% 54C P8 12W / 170W | 11400MiB / 12288MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| 0 N/A N/A 7473 C python 1862MiB |
| 0 N/A N/A 167111 C /usr/bin/python3 9002MiB |
| 0 N/A N/A 1095938 C /usr/bin/ollama 408MiB |
| 0 N/A N/A 3873555 C python3 102MiB |
+-----------------------------------------------------------------------------------------+
trevor@server:~$ nvtop
No GPU to monitor.
trevor@server:~$ nvtop --version
nvtop version 3.2.0
Same original problem here .... "No GPU to monitor"
- Ubuntu Server 24.04 LTS
- NVIDIA GeForce RTX 3060.
- NVIDIA driver 570.158.01
+-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 570.158.01 Driver Version: 570.158.01 CUDA Version: 12.8 | |-----------------------------------------+------------------------+----------------------+ | 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 RTX 3060 Off | 00000000:2D:00.0 Off | N/A | | 0% 54C P8 12W / 170W | 11400MiB / 12288MiB | 0% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 7473 C python 1862MiB | | 0 N/A N/A 167111 C /usr/bin/python3 9002MiB | | 0 N/A N/A 1095938 C /usr/bin/ollama 408MiB | | 0 N/A N/A 3873555 C python3 102MiB | +-----------------------------------------------------------------------------------------+ trevor@server:~$ nvtop No GPU to monitor. trevor@server:~$ nvtop --version nvtop version 3.2.0
https://github.com/Syllo/nvtop/issues/318#issuecomment-3038980807 This solved my problem, I completely removed Nvidia driver and only install toolkit, check my previous reply
I found another utility called "nvitop" that worked on the first try. My NVIDIA drivers are working perfectly fine on my server.
I had same issue with no gpu detection with nvtop. However if I ran it
sudo nvtop
it works? whatever I will take it.
