Graph Processing with Heterogeneous Processors

Updated: September 23, 2021

Exploring graph algorithm execution across heterogeneous processors.

Documentation

Find documentation under the wiki. For further help, contact Chris on Slack or at chrisliu@cs.ucla.edu.

Goals

• Graph generation and I/O.
• Construct BFS and SSSP kernels in C++ w/ OpenMP, CUDA, Vulkan.
• Benchmark suite for various epoch kernels.
• Heterogeneous kernels that uses both CPU and GPU variants.

Contributors

• Professor Tyler Sorensen - Principle Investigator (Summer 2021-Present)
• Christopher Liu - Part-time Contributor (Fall 2021-Present); Summer Researcher (Summer 2021)