Computing single-source shortest paths on graphs with over 8 trillion edges
Please always quote using this URN: urn:nbn:de:0297-zib-88180
- This paper introduces an implementation for solving the single-source shortest path problem on distributed-memory machines. It is tailored to power-law graphs and scales to trillions of edges. The new implementation reached 2nd and 10th place in the latest Graph500 benchmark in June 2022 and handled the largest and second-largest graphs among all participants.
Author: | Daniel RehfeldtORCiD, Katsuki FujisawaORCiD, Thorsten KochORCiD, Masahiro NakaoORCiD, Yuji ShinanoORCiD |
---|---|
Document Type: | ZIB-Report |
Date of first Publication: | 2022/11/16 |
Series (Serial Number): | ZIB-Report (22-22) |
ISSN: | 1438-0064 |