I have followed the installation steps described in this link using the package manager installation method for Ubuntu 16.04

  • I have a GeForce GTX 960M
  • gcc --version returns: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.10) 5.4.0 20160609
  • I have run sudo apt-get install linux-headers-$(uname -r) as the guide suggests
  • Downloaded the base installer from here
  • I was not able to verify the md5sum because the downloaded version is not listed on this webpage
  • I already have the nvidia 396.54 drivers installed and running, this is the output of the command nvidia-smi: nvidia-smi output
  • I navigate in the terminal to the dir where I downloaded the deb file (Downloads) and then run

sudo dpkg -i cuda-repo-ubuntu1604-9-2-local_9.2.148-1_amd64.deb

(Reading database ... 501235 files and directories currently installed.)
Preparing to unpack cuda-repo-ubuntu1604-9-2-local_9.2.148-1_amd64.deb ...
Unpacking cuda-repo-ubuntu1604-9-2-local (9.2.148-1) over (9.2.148-1) ...
Setting up cuda-repo-ubuntu1604-9-2-local (9.2.148-1) ...

sudo apt-key add /var/cuda-repo-9-2-local/7fa2af80.pub


sudo apt-get update

Hit:1 http://us.archive.ubuntu.com/ubuntu xenial InRelease
Ign:2 http://dl.google.com/linux/chrome/deb stable InRelease                                                                                                     
Get:3 http://us.archive.ubuntu.com/ubuntu xenial-updates InRelease [109 kB]                                                                                      
Hit:4 https://download.sublimetext.com apt/stable/ InRelease                                                                                                     
Get:5 http://us.archive.ubuntu.com/ubuntu xenial-backports InRelease [107 kB]                                                                                    
Hit:6 http://ppa.launchpad.net/audio-recorder/ppa/ubuntu xenial InRelease                                                                                        
Get:7 http://security.ubuntu.com/ubuntu xenial-security InRelease [107 kB]                                                                                       
Get:8 https://desktop-download.mendeley.com/download/apt stable InRelease [2,456 B]                                                                              
Hit:9 http://dl.google.com/linux/chrome/deb stable Release                                                                                                       
Hit:10 http://packages.ros.org/ros/ubuntu xenial InRelease                                                                                                       
Hit:11 http://ppa.launchpad.net/danielrichter2007/grub-customizer/ubuntu xenial InRelease                                                                  
Hit:12 http://repo.acestream.org/ubuntu trusty InRelease                                                                             
Hit:13 http://ppa.launchpad.net/graphics-drivers/ppa/ubuntu xenial InRelease                                                         
Fetched 325 kB in 0s (411 kB/s)
Reading package lists... Done
W: http://repo.acestream.org/ubuntu/dists/trusty/InRelease: Signature by key E1254F21D636B7EFDE41D2AF50E2BCF0E3805CD8 uses weak digest algorithm (SHA1)

sudo apt-get install cuda

Reading package lists... Done
Building dependency tree       
Reading state information... Done
E: Unable to locate package cuda

I think I followed all instructions very closely, I am not sure what the problem is.

  • Does anything not run -- nvcc, the graphics samples,...? Those instructions are older than the CUDA 9.2 package you installed. – ubfan1 Nov 14 '18 at 0:56
  • Well the installation never took place because the package was not found, so nvcc obviously does not run. The instructions (in the first link) are the official ones for CUDA 9.2.148 which is the version I am trying to install (CUDA 10 is not yet compatible with Pytorch) – Pedro Piacenza Nov 14 '18 at 13:45
  • Ok, so I fixed this following the instructions from this post – Pedro Piacenza Nov 14 '18 at 16:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.