Table 2 Most recent and well-known graph partitioning techniques suitable for FPGA implementation
Graph partitioning algorithm | Methodology | Programming Language | Graph partitioning | Source code | Platform | Year |
|---|---|---|---|---|---|---|
GridGraph [73] | Grid partition of edges | C++ | Store edge partition blocks on disk | Public | CPU | 2015 |
Lumos [57] | Grid partition of edges plus cross-iteration propagation values support bulk synchronous processing | C++ | Store edge partitions as blocks on disk | Public | CPU | 2020 |
FabGraph [51] | Grid partition of edges plus hash partitioning to support power law graphs | C++ | Store partition blocks on disk | not public | Multi-FPGA | 2019 |
PowerGraph [31] | Vertex-cut partitioning | C++, Java, Scala | Partitioning during Runtime | Public | CPU | 2013 |
Graphchi [37] | Shard-interval partitioning plus sorting, asynchronous processing | HLS | Partitioning during Runtime | Public | Multi-FPGA | 2021 |
ThunderGP [16] | Vertex-cut partitioning | HLS-C/C++ | Partitioning during Runtime | Public | single-FPGA | 2021 |
Foregraph [19] | Shard-interval | HDL | Partitioning during Runtime | not Public | Multi-FPGA | 2017 |