pub async fn gpu_bfs_paged(
device: &GpuDevice,
graph: &CsrGraph,
source: NodeId,
) -> Result<GpuBfsResult>Expand description
Run GPU breadth-first search over a graph that may exceed VRAM.
When the graph fits in GPU memory this delegates to the in-VRAM
super::gpu_bfs; otherwise it tiles the graph via PagingCoordinator and
processes each tile in turn.
ยงErrors
Returns an error if GPU paging setup or any tile traversal fails.