SEP-Graph-source-code

This release includes the source code of the PPoPP'19 paper "SEP-Graph: Finding Shortest Execution Paths for Graph Processing under a Hybrid Framework on GPU ". The artifact evaluation is excluded in this release but is put into the release of SEP-Graph-AE separately.

SEP-Graph is an open-source framework for graph processing on GPU. It supports the graph algorithms to be implemented as the combinations of Sync. or Async execution mode, Pull or Push communication method, and Data-driven or Topology-driven vertex traversing mechanism. At runtime, SEP-Graph will try to find the shortest execution path of a graph algorithm in each iteration.

This work is designed, implemented, and maintained by the High Performance Computing and Software Lab, led by Professor Xiaodong Zhang from the Department of Computer Science and Engineering, Ohio State University.

SEP-Graph-source-code

This release includes the source code of the PPoPP'19 paper "SEP-Graph: Finding Shortest Execution Paths for Graph Processing under a Hybrid Framework on GPU ". The artifact evaluation is excluded in this release but is put into the release of SEP-Graph-AE separately.

SEP-Graph is an open-source framework for graph processing on GPU. It supports the graph algorithms to be implemented as the combinations of Sync. or Async execution mode, Pull or Push communication method, and Data-driven or Topology-driven vertex traversing mechanism. At runtime, SEP-Graph will try to find the shortest execution path of a graph algorithm in each iteration.

This work is designed, implemented, and maintained by the High Performance Computing and Software Lab, led by Professor Xiaodong Zhang from the Department of Computer Science and Engineering, Ohio State University.