Difference between revisions of "GPGPU"

From RedwoodCenter
Jump to navigationJump to search
 
Line 1: Line 1:
 
This wiki page contains information on how to use the graphics processing units (GPU) on the graphics card for general-purpose computing. For general information on the cluster, please see [[cluster|here]]
 
This wiki page contains information on how to use the graphics processing units (GPU) on the graphics card for general-purpose computing. For general information on the cluster, please see [[cluster|here]]
 +
 +
== OpenCL ==
 +
 +
'''note: OpenCL doesn't work yet. Here is what I tried so far'''
 +
 +
The OpenCL applications in the NVIDIA GPU Computing SDK require a GPU with CUDA Compute Architecture to run properly. In order to use OpenCL, you have to load the CUDA environment:
 +
 +
  module load cuda
 +
 +
Install version 2.3 of the NVIDIA GPU Computing SDK by executing the followin command
 +
 +
  bash /clusterfs/cortex/software/opencl-2.3/src/gpucomputingsdk_2.3a_linux.run
 +
 +
It seems that you need gcc 4.3 in order to compile SDK. Change the following three lines in ~/NVIDIA_GPU_Computing_SDK/OpenCL/common/common_opencl.mk from
 +
 +
  CXX        := g++
 +
  CC        := gcc
 +
  LINK      := g++ -fPIC
 +
 +
to
 +
 +
  CXX        := g++43
 +
  CC        := gcc43
 +
  LINK      := g++43 -fPIC
 +
 +
You can compile all the code examples by running (fails at the moment):
 +
 +
  export LIBRARY_PATH=$LIBRARY_PATH:~/NVIDIA_GPU_Computing_SDK/OpenCL/common/lib/Linux64
 +
  export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:~/NVIDIA_GPU_Computing_SDK/OpenCL/common/lib/Linux64
 +
 +
  cd ~/NVIDIA_GPU_Computing_SDK/OpenCL
 +
  make
 +
 +
The compiled examples can be found in the directory
 +
 +
  ~/NVIDIA_GPU_Computing_SDK/OpenCL/bin/linux/release

Revision as of 13:12, 13 September 2009

This wiki page contains information on how to use the graphics processing units (GPU) on the graphics card for general-purpose computing. For general information on the cluster, please see here

OpenCL

note: OpenCL doesn't work yet. Here is what I tried so far

The OpenCL applications in the NVIDIA GPU Computing SDK require a GPU with CUDA Compute Architecture to run properly. In order to use OpenCL, you have to load the CUDA environment:

 module load cuda

Install version 2.3 of the NVIDIA GPU Computing SDK by executing the followin command

 bash /clusterfs/cortex/software/opencl-2.3/src/gpucomputingsdk_2.3a_linux.run

It seems that you need gcc 4.3 in order to compile SDK. Change the following three lines in ~/NVIDIA_GPU_Computing_SDK/OpenCL/common/common_opencl.mk from

 CXX        := g++
 CC         := gcc
 LINK       := g++ -fPIC

to

 CXX        := g++43
 CC         := gcc43
 LINK       := g++43 -fPIC

You can compile all the code examples by running (fails at the moment):

 export LIBRARY_PATH=$LIBRARY_PATH:~/NVIDIA_GPU_Computing_SDK/OpenCL/common/lib/Linux64
 export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:~/NVIDIA_GPU_Computing_SDK/OpenCL/common/lib/Linux64
 cd ~/NVIDIA_GPU_Computing_SDK/OpenCL
 make

The compiled examples can be found in the directory

 ~/NVIDIA_GPU_Computing_SDK/OpenCL/bin/linux/release