AlgorandSIMULATION v1.0
Sequential Engines(17)
Parallel Engines
Parallel BFS
parallelPRAM-CREWA parallelized version of graph exploration that processes all frontier nodes concurrently, writing newly visited nodes to the next frontier in parallel.
SIMULATION STEP PROGRESSStep 1 / 1
SPEED (50%)
CORES (4)
Code Workspace
1procedure ParallelBFS(G, start_node)
2 level[start_node] := 0
3 frontier := {start_node}
4 while frontier is not empty do
5 next_frontier := empty set
6 for each u in frontier in parallel do
7 for each neighbor v of u in parallel do
8 if level[v] is unassigned then
9 level[v] := level[u] + 1
10 add v to next_frontier (concurrent write/sync)
11 end if
12 end parallel-for
13 end parallel-for
14 frontier := next_frontier
15 barrier()
16 end while
17end procedure
Currently highlights executing branch
RUNTIME COST METRICS
Comparisons0
Memory Writes0
Memory Reads0
Elements Swapped0
Core Capacity:4 Cores
Total Work Done:5 operations
Critical Path Span:2 levels
Observed Speedup:2.5x
Efficiency Bound:63%
PROCESSOR ALLOCATION EFFICIENCY
THEORETICAL ANALYSIS & RESEARCH
Conceptual Focus: A parallelized version of graph exploration that processes all frontier nodes concurrently, writing newly visited nodes to the next frontier in parallel.
Real-world Industrial Applications:
- Social network analytics (analyzing connection components)
- Web crawlers scanning multiple pages in parallel