Parallel: General Purpose Graphics Processing Units, Part 1
SESSION: Parallel: GPGPU, Part 1
EVENT TYPE: Communities, Education
TIME: 2:00PM - 3:00PM
Speaker(s):Charlie Peck, Henry Neeman, Tom Murphy, Dan Ernst
ABSTRACT: This session gives an introduction to NVIDIA's CUDA, a platform for using Graphics Processing Units (GPUs) as powerful parallel accelerators for general purpose applications. The prerequisites are at least one semester of programming experience in C, C++ and/or Fortran, recently, and at least one semester of experience with any Unix-like operating system (including but not limited to Linux), recently. Topics include: an introduction to GPU architectures and programming models (Thread/Block/Grid/Kernel concepts); the basics of getting data to and from a GPU using the CUDA API; hands-on experience with first CUDA Kernels.