Don't have Telegram yet? Try it now!
https://hgpu.org/?p=7285
Better speedups using simpler parallel programming for graph connectivity and biconnectivity