GPU Computing Workshop
Garland Durham, Cal Poly
Getting ready
- Workshop participants should bring a laptop computer. A large part of
the workshop will involve hands-on exercises. Access to Amazon EC2 GPU servers will be
provided. To access these, please be sure that SSH and SFTP clients are
installed on your laptop (see SSH notes).
- If you have an Nvidia GPU on your laptop and would like to use it,
install the CUDA driver and SDK from the CUDA website. You
will also need a C/C++ compiler (see Nvidia installation notes for
details).
- If you want to use your own GPU, it may be useful to install some
additional software (this is not necessary for using the provided Amazon
EC2 GPU servers).
- Matlab with parallel toolbox, if available.
- Python, with SciPy and PyCUDA packages.
- Accelereyes Arrayfire library for CUDA.
Workshop materials
Additional resources
Here are a few reference materials that may be useful: