Parallel Computing with GPUs

Parallel Computing with GPUs (COM4521/COM6521) is a 4th/MSc level academic module within the school of Computer Science, led by Robert Chisholm from the RSE team.

The module provides an introduction to utilising CPU and GPU parallel computing through OpenMP and CUDA.

Learning Objectives:

• Compare and contrast parallel computing architectures
• Implement programs for GPUs and multicore architectures
• Apply benchmarking and profiling to GPU programs to understand performance
• Identify and address limiting factors and apply optimisation to improve code performance

Research students and staff wishing to sit the module (either assessed or un-assessed) should contact com-teaching@sheffield.ac.uk. This may not be permitted where the module’s capacity has been reached by taught students.

Due to current university policy teaching materials are now held on Blackboard, however an archived course webpage and materials can be found here.

For further information please contact robert.chisholm@sheffield.ac.uk.