Strong Scaling Achieves 15.2× Speedup for Dynamic Graph Updates with Multi-Threaded Efficiency

Too Long; Didn't Read

Dynamic Frontier PageRank demonstrates robust strong-scaling behavior for batch updates, achieving an average speedup of 10.3× with 16 threads and 15.2× with 64 threads. It gains a 1.8× performance boost with every doubling of threads but faces NUMA limitations at higher thread counts. The method efficiently handles dynamic graph updates, scaling well across datasets and graph sizes.