GPU Computing

CSE101 ,CSE102

This course will introduce parallel computing paradigms with focus on GPU programming to harness the massively parallel GPU architecture in solving computationally demanding tasks. The NVIDIA CUDA and industry standard OpenCL frameworks will be introduced and used with most of the labs. This is a project based course where the students will work on scientific computational problems.

1. Students are able to understand concepts behind parallel computing.
2. Students are able to understand parallel computing paradigms, GPU architecture and GPGPU development frameworks (CUDA, OpenCL, and GLSL).
3. Students are able to analyse an algorithms to provide parallel solutions to computationally challenging problems.
4. Students are able to implement such solutions on GPU using CUDA, and show effectiveness of the GPU based solutions using standard benchmarks and tools.

Course Offering