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
6 +1 main
iters total
500M
71428571/stream
elapsed
1157.3 ms
total CPU used
7233.15 ms
speedup
6.25×
vs serial
efficiency
89.3%
of 7× ideal
stream spawn ms spawned@ work start@ work end@ work ms reap wait ms
0 (main) 0 13.97 13.98 962.76 948.78 0
1 2.232 2.25 15.04 1141.77 1126.73 179.18
2 1.826 4.1 18.35 1118.21 1099.86 158.4
3 1.807 5.92 23.31 986.62 963.31 29.33
4 4.637 10.57 44.35 993.14 948.79 35.76
5 1.697 12.28 44.35 1154.11 1109.76 191.51
6 1.658 13.96 44.35 1080.27 1035.92 117.65
main
w1
w2
w3
w4
w5
w6
    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.