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
100M
11111111/stream
elapsed
271.51 ms
total CPU used
1810.88 ms
speedup
6.67×
vs serial
efficiency
74.1%
of 9× ideal
stream spawn ms spawned@ work start@ work end@ work ms reap wait ms
0 (main) 0 28.61 28.61 266.61 238 0
1 2.22 2.24 16.89 235.19 218.3 0.16
2 1.954 4.21 34.29 226.88 192.59 0.23
3 1.818 6.05 24.64 203.38 178.74 0.24
4 1.806 7.88 36.87 260.44 223.57 0.26
5 2.306 10.2 42.21 246.6 204.39 0.27
6 2.05 12.27 42.56 237.13 194.57 0.29
7 1.823 14.11 70.4 231.51 161.11 0.33
8 14.465 28.59 68.81 268.42 199.61 1.92
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.