How to use Cuda on my Computer

[apollo000104]

Hello, Everybody.
I have nvidia geForce GTX 750Ti on my computer.
I can see

like this, with "!nvidia-smi" command.
But the result of this prompt is "False"
\torch.cuda.is_available()\

Why? How can I overcome this problem?
help me please.🙏

[otrturn]

Have you installed torch with cuda? I.e. check whether during the installation you used someting like pip install torch==2.0.1+cpu. In this case you may have a GPU on your system, but torch without cuda-support.

[apollo000104]

yes, I can see torch==2.0.1+cpu after the command torch.version.
So How can I do for new installation of torch with cuda????

[otrturn]

Check the site https://pytorch.org/get-started/locally/. You may have to uninstall torch.

[apollo000104]

Got it. I will try it.
Thanks

[apollo000104]

Hello,Thanks very much for your help. I have tried, but it is not working. I have uninstalled torch and newly install torch cuda 118 But it is not still working

…

On Wed, Aug 30, 2023 at 12:26 AM Ralf Scholler ***@***.***> wrote: Check the sitehttps://pytorch.org/get-started/locally/ — Reply to this email directly, view it on GitHub <#617 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A73T322HCVKNUSYLG4IBR33XXYC2LANCNFSM6AAAAAA4DDBQKM> . You are receiving this because you authored the thread.Message ID: <mrdbourke/pytorch-deep-learning/repo-discussions/617/comments/6854595@ github.com>

[otrturn]

Hello Apollo,

my workstation with an NVIDIA GPU has arrived today. So, I must go through the same ordeal. I will document everything and will send it to you. How does that sound?

Ciao

Ralf

[apollo000104]

I wonder If you do so.
very thanks

[otrturn]

Hello Apollo,

I sure will. It is not an easy task to get the GPU running with jupyter, believe me. Ubuntu was pre-installed, the GPU was connected and worked. Then I tried to install the CUDA toolkit and the cuDNN Library. The result was that the workstation crashed as it mixed up drivers of different versions. As soon as I have better news, I update you.

[mrdbourke]

I've had similar issues setting up a deep learning PC in the past.

I've collected a bunch of notes here on how I got mine setup.

In short, I had to remove all existing NVIDIA drivers and reinstall them.

These notes are rough and not a 100% working guide but they've got resources that helped me: https://mrdbourke.notion.site/Installing-CUDA-on-a-new-GPU-175916438b814525b8b161ac0c787667?pvs=4

[apollo000104]

Ok, thanks for the update.

…

On Fri, Sep 1, 2023 at 12:00 AM Ralf Scholler ***@***.***> wrote: Hello Apollo, I sure will. It is not an easy task to get the GPU running with jupyter, believe me. Ubuntu was pre-installed, the GPU was connected and worked. Then I tried to install the CUDA toolkit and the cuDNN Library. The result was that the workstation crashed as it mixed up drivers of different versions. As soon as I have better news, I update you. — Reply to this email directly, view it on GitHub <#617 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A73T3275ORVWNCSPSJ5CVADXYCRKRANCNFSM6AAAAAA4DDBQKM> . You are receiving this because you authored the thread.Message ID: <mrdbourke/pytorch-deep-learning/repo-discussions/617/comments/6876836@ github.com>

[apollo000104]

Awesome, thanks!

…

On Thu, Sep 7, 2023 at 9:28 AM Daniel Bourke ***@***.***> wrote: I've had similar issues setting up a deep learning PC in the past. I've collected a bunch of notes here on how I got mine setup. In short, I had to *remove* all existing NVIDIA drivers and reinstall them. These notes are rough and not a 100% working guide but they've got resources that helped me: https://mrdbourke.notion.site/Installing-CUDA-on-a-new-GPU-175916438b814525b8b161ac0c787667?pvs=4 — Reply to this email directly, view it on GitHub <#617 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A73T323HETLE2N5KXOHZQPDXZEIJVANCNFSM6AAAAAA4DDBQKM> . You are receiving this because you authored the thread.Message ID: <mrdbourke/pytorch-deep-learning/repo-discussions/617/comments/6931761@ github.com>

[otrturn]

Has the issue been solved? I have been out of business for a while.

[apollo000104]

Yes, everything is fine. Thanks.

…

On Thu, Sep 21, 2023 at 5:41 PM Ralf Scholler ***@***.***> wrote: Has the issue been solved? I have been out of business for a while. — Reply to this email directly, view it on GitHub <#617 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A73T323KAIKCDZO7NRIXWPTX3P4VBANCNFSM6AAAAAA4DDBQKM> . You are receiving this because you authored the thread.Message ID: <mrdbourke/pytorch-deep-learning/repo-discussions/617/comments/7067678@ github.com>

[apollo000104]

How are things going? doing well? If you have any problem on your work, call me plz. I will help you

…

On Thu, Sep 21, 2023 at 10:52 PM Andrew leon ***@***.***> wrote: Yes, everything is fine. Thanks. On Thu, Sep 21, 2023 at 5:41 PM Ralf Scholler ***@***.***> wrote: > Has the issue been solved? I have been out of business for a while. > > — > Reply to this email directly, view it on GitHub > <#617 (reply in thread)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/A73T323KAIKCDZO7NRIXWPTX3P4VBANCNFSM6AAAAAA4DDBQKM> > . > You are receiving this because you authored the thread.Message ID: > <mrdbourke/pytorch-deep-learning/repo-discussions/617/comments/7067678@ > github.com> >

[oraclesoon]

I am using Asus VivoBook X540UBR (date: 01/08/2020) which comes with entry level NVIDIA GeForce MX110.

2 years ago, I abandoned Windows 10 and re-installed the notebook with Linux Mint 21.2 Cinnamon.

To my surprise, Linux has properly installed and maintained NVIDIA drivers via
$ sudo apt update
$ sudo apt dist-upgrade

I don't need to explicitly install Cuda toolkit from https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&Distribution=Ubuntu&target_version=20.04&target_type=deb_local

I simply install Cuda toolkit as hinted by apt via
$ sudo apt install nvidia-cuda-toolkit

Then, my GPU environment is ready.

To check the Cuda version in Ubuntu:

$ nvidia-smi

Sat Mar 16 21:16:35 2024       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.161.07             Driver Version: 535.161.07   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 MX110           Off | 00000000:01:00.0 Off |                  N/A |
| N/A   38C    P8              N/A / 200W |      6MiB /  2048MiB |      0%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+

To check whether Cuda toolkit is installed:

$ 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

I follow https://github.com/mrdbourke/pytorch-deep-learning/blob/main/SETUP.md#tk---2-getting-setup-locally-linux-version.

import torch
print(torch.cuda.is_available())
print(torch.__version__)

True
1.10.0