Instaling CUDA on Backtrack Unknown rwxr-xr-x 0 21:54

Title	Instaling CUDA on Backtrack
Permission	rw-r--r--
Author	Unknown
Date and Time	21:54
Category
Share

If you want your CPU performance increase, you can try CUDA. You must install CUDA on Backtrack. Installing CUDA on Backtrack can increase your CPU performance. CUDA is a parallel computing platform and programming model. It's created by NVIDIA so you must have NVIDIA graphic card if you want use CUDA. NVIDIA GPUs implement the CUDA architecture and programming model. CUDA increases the CPU computing performance by harnessing the power of GPU. Here is the CUDA installation on Backtrack but if you use another Unix OS, Ubuntu for example, you can install cuda Ubuntu too.

First you must prepare Backtrack kernel source for NVIDIA driver installation
prepare-kernel-sources

Then change your directory to /usr/src/linux and copy all file on include/generated/ to include/linux/
cd /usr/src/linux
cp -rf include/generated/* include/linux/

Now download NVIDIA drivers. The drivers according to your CPU architecture.
32 bit: http://developer.download.nvidia.com/compute/cuda/4_0_rc2/drivers/devdriver_4.0_linux_32_270.40.run

64 bit: http://developer.download.nvidia.com/compute/cuda/4_0_rc2/drivers/devdriver_4.0_linux_64_270.40.run

Run the Nvidia driver installer, but you must log out from X session first. Then, download the CUDA toolkit. The toolkit according to your CPU architecture.
32 bit
http://www.nvidia.com/object/thankyou.html?url=/compute/cuda/4_0_rc2/toolkit/cudatoolkit_4.0.13_linux_32_ubuntu10.10.run

64 bit
http://www.nvidia.com/object/thankyou.html?url=/compute/cuda/4_0_rc2/toolkit/cudatoolkit_4.0.13_linux_64_ubuntu10.10.run

You need to configure the Backtrack environment variables and ideally CUDA was installed in /opt. This is for nvcc command, because if you don't configure the environment variables the nvcc command won't work. You can do this by add this lines in your /root/.bashrc file.
PATH=$PATH:/opt/cuda/bin
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/cuda/lib
export PATH
export LD_LIBRARY_PATH

Then run this command
source /root/.bashrc
ldconfig

Now check that nvcc compiler was installed with run this two command
which nvcc
nvcc -V

After Nvidia driver and CUDA toolkit are completely installed, we can test with pyrit, a GPU powered tool.
svn checkout http://pyrit.googlecode.com/svn/trunk/ pyrit
cd pyrit/pyrit && python setup.py build && python setup.py install
cd ../../
cd pyrit/cpyrit_cuda && python setup.py build && python setup.py install

Then run a benchmark
pyrit benchmark

Ok that's all. the CUDA installation was completed. Next time I'll post about install CUDA Ubuntu. Installing CUDA on Ubuntu will be easy like this :-)