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
8 +1 main
iters total
500M
55555555/stream
elapsed
1176.23 ms
total CPU used
8933.43 ms
speedup
7.59×
vs serial
efficiency
84.3%
of 9× ideal
stream spawn ms spawned@ work start@ work end@ work ms reap wait ms
0 (main) 0 20.08 20.09 847.74 827.65 0
1 2.14 2.16 21.02 1157.69 1136.67 310.08
2 1.923 4.1 25.76 1043.84 1018.08 199.21
3 1.817 5.94 26.32 965.68 939.36 121.04
4 1.745 7.7 35.99 977.76 941.77 130.12
5 1.621 9.34 52.78 1172.9 1120.12 325.29
6 1.638 11 43.63 1015.04 971.41 167.41
7 4.338 15.35 53.63 951.65 898.02 105.78
8 4.695 20.07 63.66 1144.01 1080.35 296.41
main
w1
w2
w3
w4
w5
w6
w7
w8
    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.