CPU scaling benchmark

W=0 W=1 W=2 W=3 W=4 W=6 W=8 W=9 W=10 W=12 W=16 iters: 100M 500M 1B 2B
workers
9 +1 main
iters total
500M
50000000/stream
elapsed
1167.84 ms
total CPU used
9893.36 ms
speedup
8.47×
vs serial
efficiency
84.7%
of 10× ideal
stream spawn ms spawned@ work start@ work end@ work ms reap wait ms
0 (main) 0 32.48 32.49 893.35 860.86 0
1 1.847 1.86 15.81 1112.15 1096.34 221.22
2 1.415 3.3 22.62 1097.04 1074.42 208.9
3 2.665 5.98 23.34 957.4 934.06 66.85
4 2.567 8.57 44.7 1119.22 1074.52 226
5 3.159 11.75 40.38 1164.8 1124.42 271.59
6 1.374 13.14 48.58 755.07 706.49 0.13
7 9.997 23.16 64.6 1165.2 1100.6 274.39
8 3.15 26.32 54.6 975.5 920.9 87.17
9 6.127 32.47 74.67 1075.42 1000.75 187.11
main
w1
w2
w3
w4
w5
w6
w7
w8
w9
    fork+handshake      CPU work      parent reap wait
what this measures
Each stream runs a tight integer LCG loop — working set is one CPU register, no memory access, no shared data. Speedup = sum(stream CPU time) / wall-clock elapsed. Efficiency = speedup / (workers+1). 100% efficiency means perfect linear scaling; less than 100% is the cost of serial fork setup, reap tail, SMT/core contention.