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
1157.15 ms
total CPU used
9114.52 ms
speedup
7.88×
vs serial
efficiency
87.6%
of 9× ideal
stream spawn ms spawned@ work start@ work end@ work ms reap wait ms
0 (main) 0 13.93 13.96 1028.5 1014.54 0
1 1.996 2.01 16.68 1000.35 983.67 0.14
2 1.933 3.97 55.31 1035.52 980.21 8.18
3 1.681 5.67 34.25 1016.02 981.77 0.2
4 1.749 7.44 44.94 1154.29 1109.35 125.89
5 1.622 9.09 50.47 1150.05 1099.58 121.69
6 1.537 10.64 62.28 908.81 846.53 8.14
7 1.618 12.28 72.26 1129.07 1056.81 100.71
8 1.607 13.91 58.98 1101.04 1042.06 72.65
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.