nvidia

Install drivers

Download driver from:

• https://www.nvidia.com/content/DriverDownload-March2009/confirmation.php?url=/XFree86/Linux-x86_64/410.78/NVIDIA-Linux-x86_64-410.78.run&lang=us&type=TITAN
• or from https://www.nvidia.com for your Card

sudo bash NVIDIA-Linux-x86_64-410.78.run

need to disable nouveau, run script will detect and disable by adding to blacklist.

update initramfs to apply.

sudo update-initramfs -u

disable graphic ui

sudo systemctl set-default multi-user.target

reboot. screen may go to black empty. login with terminal.

re-run

sudo bash NVIDIA-Linux-x86_64-410.78.run

enable gui

sudo systemctl set-default graphical.target

reboot again.

cuda tool kit

https://developer.nvidia.com/cuda-downloads

cuda tool kit 10.1

(2019-04-10)

wget https://developer.nvidia.com/compute/cuda/10.1/Prod/local_installers/cuda_10.1.105_418.39_linux.run

sudo sh cuda_10.1.105_418.39_linux.run

Result

• There is another attempt to install the driver
• As it is loaded and used, this fails
• cuda 10.1 installation is success

cuda tool kit 10.0

• https://developer.nvidia.com/cuda-downloads
• need nouveau removed
• nouveau seems to make problems with some hardware and keeps segmentation fault, logs in dmesg
• this keeps booting and shutdown pretty long... :(

sudo ./cuda_10.0.130_410.48_linux.run

message

Please make sure that

• PATH includes /usr/local/cuda-10.0/bin
• LD_LIBRARY_PATH includes /usr/local/cuda-10.0/lib64, or, add /usr/local/cuda-10.0/lib64 to /etc/ld.so.conf and run ldconfig as root

To uninstall the CUDA Toolkit, run the uninstall script in /usr/local/cuda-10.0/bin To uninstall the NVIDIA Driver, run nvidia-uninstall

Please see CUDA_Installation_Guide_Linux.pdf in /usr/local/cuda-10.0/doc/pdf for detailed information on setting up CUDA.

Logfile is /tmp/cuda_install_7988.log

A test

 ./matrixMulCUBLAS
[Matrix Multiply CUBLAS] - Starting...
GPU Device 0: "GeForce GTX 1050 Ti" with compute capability 6.1

GPU Device 0: "GeForce GTX 1050 Ti" with compute capability 6.1

MatrixA(640,480), MatrixB(480,320), MatrixC(640,320)
Computing result using CUBLAS...done.
Performance= 1201.25 GFlop/s, Time= 0.164 msec, Size= 196608000 Ops
Computing result using host CPU...done.
Comparing CUBLAS Matrix Multiply with CPU results: PASS

NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled.

cuDNN

sudo tar xvf cudnn-10.0-linux-x64-v7.4.1.5.tgz -C /usr/local

add /usr/local/cuda/lib64 to /etc/ld.so.conf/cuda.conf and

sudo ldconfig

Test

cp -R /usr/src/cudnn_samples_v7/mnistCUDNN ~/work/.
cd ~/work/mnistCUDNN 
make clean
make
./mnistCUDNN

cuda tool kit 9.1

sudo apt install nvidia-cuda-toolkit