This repository provides set up instructions and the exerises for the CppCon parallel programing class.

The majority of the exercises will require a only standard C++17 compliant compiler, however if for any reason you cannot use a C++17 compiler, using online compilers such as https://godbolt.org/ and https://wandbox.org/ will suffice.

Some later exercises, those which involve programming for the GPU will have some further dependencies. You will need to install OpenCL drivers for the GPU on your laptop and ComputeCpp SYCL (see instructions below). If you do not have a GPU on your laptop or cannot find suitable OpenCL drivers for your GPU then the CPU will suffice. If you have trouble setting this up or for any reason are unable to install the requirements on your laptop we are providing a docker image which will provide OpenCL drivers for Intel CPU and ComputeCpp SYCL (see instructions below).

• You will need the OpenCL drivers for your GPU:

  • Intel OpenCL drivers: https://software.intel.com/en-us/articles/opencl-drivers.
  • AMD OpenCL drivers: https://www.amd.com/en/support.
  • NVidia OpenCL drivers: https://developer.nvidia.com/cuda-toolkit-32-downloads

• You can download ComputeCpp SYCL and find many useful resources at https://developer.codeplay.com/.

• For help getting setup and verifying your installation see https://developer.codeplay.com/computecppce/latest/getting-started-guide.

• In general ComputeCpp supports Windows 7/10 and Ubuntu 16.04/18.04, and we support Intel CPU/GPU, AMD GPU and have experimental support for NVidia GPU on Ubunutu. For a full list of the supported platforms see https://developer.codeplay.com/computecppce/latest/supported-platforms.

• Note that ComputeCpp currently only supports the earlier Radeon Software Crimson Edition AMD drivers.

• Install Docker CE from https://docs.docker.com/install/.

• Pull the computecpp_ubunutu1604 docker image:

  docker pull aerialmantis/computecpp_ubuntu1604

• Run the computecpp_ubuntu1604 docker image:

  docker run --rm -it aerialmantis/computecpp_ubuntu1604

• Once you've got OpenCL and ComputeCpp SYCL setup or you have the docker image setup you can clone this repository to verify your setup and prepare yourself for the class.

• Pull this repository:

  git clone https://github.com/AerialMantis/cppcon2018-parallelism-class.git

• Create a build directory:

  cd cppcon2018-parallelism-class/ mkdir build cd build

• Run CMake to configure solution (local setup):

  cmake ../ -G[Generator] -DCMAKE_BUILD_TYPE=[Debug/Release]

• Note that if you are using an NVidia GPU, in order to use the experimental ComputeCpp SYCL support you must include the following in the above cmake command:

  -DCOMPUTECPP_BITCODE=ptx64

• Run CMake to configure solution (docker image):

  cmake ../ -GNinja -DCMAKE_BUILD_TYPE=Debug -DOpenCL_LIBRARY=${OCL_LIB}/libOpenCL.so -DOpenCL_INCLUDE_DIR=${OCL_INC}

• Build your solution:

  cmake --build

• Verify your setup by running the hello world example:

  ./examples/hello_world