How To Install Nvidia CUDA 9.0 Toolkit on Ubuntu 18.04

If the purpose of installing the CUDA toolkit 9.0 on Ubuntu 18.04 is purely to use tensorflow-gpu, I strongly advise you to use the Docker method documented here, as you get better hardware and code isolation and easy portability to the cloud later.
TensorFlow Docker Installation incorrectly indicates that the host machine needs the CUDA toolkit and cuDNN libraries to be installed on the host machine. This is not true. I have reached out Tensorflow community to correct this.

How To

If you need to install CUDA 9.0 toolkit on Ubuntu 18.04, here are the instructions.

Uninstall Nvidia Drivers

Uninstall any Nvidia drivers. To uninstall do:
sudo apt-get remove --purge nvidia-*
sudo apt-get remove --purge libnvidia-* This will avoid any versioning conflicts. CUDA Toolkit 9.0 and 9.1 requires nvidia-390_390.30 drivers. 

Install CUDA Toolkit 9.0 (using Ubuntu 17.04 repository meta-data)

The CUDA Toolkit is available at https://developer.nvidia.com/cuda-toolkit. You will have to navigate to the archives to find the install for CUDA 9.0 repository meta-data file for Ubuntu 17.04. 

Direct link:
https://developer.nvidia.com/cuda-90-download-archive?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1704&target_type=debnetwork

The required steps are:
1) Download the network repository meta-data. Navigate to the folder and install
2) Installing the CUDA public GPG key
3) Update the apt repository cache
4) Install CUDA

To perform the steps:
sudo apt-get -i cuda-repo-ubuntu1704_9.0.176-1_amd64.deb
sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1704/x86_64/7fa2af80.pub
sudo apt-get update
sudo apt-get install cuda-9-0
Be sure to explicitly type out the desired version. If you do not it will install 9.1! I don't know why it does this, but it happened.

Now, the install process can take a bit of time. If all goes well you will not have an installation error.
If you come across an error use this to fix the install:
sudo apt-get install -f
If you get:
Unpacking nvidia-390 (390.30-0ubuntu1) ...
dpkg: error processing archive /var/cache/apt/archives/nvidia-390_390.30-0ubuntu1_amd64.deb (--unpack):
 trying to overwrite '/usr/lib/x86_64-linux-gnu/libGLX_indirect.so.0', which is also in package libglx-mesa0:amd64 18.0.0~rc5-1ubuntu1
Errors were encountered while processing:
 /var/cache/apt/archives/nvidia-390_390.30-0ubuntu1_amd64.deb
E: Sub-process /usr/bin/dpkg returned an error code (1)
To fix this, check if the libGX_indirect.so.0 file exists. If it exists make a backup:
sudo mv /usr/lib/x86_64-linux-gnu/libGLX_indirect.so.0 /usr/lib/x86_64-linux-gnu/libGLX_indirect.so.0.old
The weird thing for me is the file didn't exist! Nonetheless, this command will install force the install of the drivers:
dpkg -i --force-overwrite /var/cache/apt/archives/nvidia-390_390.30-0ubuntu1_amd64.deb
Run the install again, and succeed!
sudo apt-get install cuda-9-0

Install cuDNN 7.1.4 (using Ubuntu 16.04 repository meta-data)

Go to: https://developer.nvidia.com/cudnn and register a free Nvidia Developer Account to get access. Use the latest 7.1.4 version of cuDNN corresponding to CUDA toolkit 9.0.

Select the Ubuntu 16.04 Deb files:

cuDNN v7.1.4 Runtime Library for Ubuntu16.04 (Deb)
cuDNN v7.1.4 Developer Library for Ubuntu16.04 (Deb)
cuDNN v7.1.4 Code Samples and User Guide for Ubuntu16.04 (Deb)

Install all three packages by:
sudo dpkg -i ./libcudnn7_7.1.4.18-1+cuda9.2_amd64.deb
sudo dpkg -i ./libcudnn7-dev_7.1.4.18-1+cuda9.0_amd64.deb
sudo dpkg -i ./libcudnn7-doc_7.1.4.18-1+cuda9.0_amd64.deb

Comments

Post a Comment

Popular posts from this blog

How Salesforce uses AWS to Improve The Support Call Experience

Image
In April I wrote about how moving infrastructure to  the cloud is the start of a paradigm shift, and that building intelligence products and services is the next step. Today, I highlight an example of this presented by Salesforce and AWS . Youtube video link:  https://www.youtube.com/watch?v=TEiDEFFZjNw The key AWS services discussed in the video are: Amazon Lex Amazon Lex is a chatbot platform. It uses speech recognition and language understanding technology to analyze text and speech. Amazon Lex can be trained to recognize a users intent and carry out a conversation to collect the necessary data. Amazon Polly Amazon Polly is a text-to-speech service. You supply a string of text and Amazon Polly will return an audio file in a variety of different voices. To learn more about the features see  https://aws.amazon.com/polly/features/ . To see the currently supported languages and voices available for those languages see  https://docs.aws.amazon...
Read more

Apple Pay, Android Pay, contactless credit cards, is it safe?

Image
If you live in Canada and own a credit card you are now probably accustomed to tapping your credit card to pay for your small purchases. Canadian banks started issuing contactless credit cards in 2010, and most vendors now have pay terminals that support "Tap". Visa officially calls the contact-less credit card feature payWave  and MasterCard calls it  Tap & Go . For majority of Americans this great convenient form of payment came in 2014 with  Apple Pay  or Android Pay . Canadian banks are starting to enabling Apple Pay this  June . Apple Pay, and Android Pay use NFC. Early versions of the contactless credit cards used RFID and was a hot topic in 2012 as hackers demonstrated how the can read the data  just by standing near you with a smartphone. Since then Visa payWave and MasterCard Tap &Go have moved to using NFC. What is NFC and how is it different from RFID? NFC stands for Near Field Communication and is related to b...
Read more

Failed CUDA Toolkit Install? Ubuntu 18.04 stuck on boot of Gnome Display Manager?

Image
I have been attempting to get TensorFlow GPU running on Ubuntu 18.04. The system requirements are simple:  TensorFlow GPU NVIDIA requirements . But I discovered the CUDA toolkit can result in some messy installation dependencies on specific versions of the Nvidia drivers leaving you with failed package installations such as: Errors were encountered while processing: /tmp/apt-dpkg-install-5dMwo8/100-nvidia-390_390.30-0ubuntu1_amd64.deb E: Sub-process /usr/bin/dpkg returned an error code (1) What's worse if you reboot your computer you are met with a computer that is stuck booting, the system not able to progress past loading the Gnome Display Manager. Is my computer completely messed up? If you are stuck at boot up sequence access a tty terminal by pressing  ctrl + alt + F2 . You will be prompted to login in. Do it. Next, type: sudo apt-get install -f The system will list dependencies that were not installed correctly. In my case, my installed ...
Read more