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
4 +1 main
iters total
500M
100000000/stream
elapsed
1259.06 ms
total CPU used
5357.55 ms
speedup
4.26×
vs serial
efficiency
85.2%
of 5× ideal
stream spawn ms spawned@ work start@ work end@ work ms reap wait ms
0 (main) 0 7.49 7.51 1216.83 1209.32 0
1 1.905 1.92 18.05 997.39 979.34 0.12
2 1.533 3.48 20.1 939.4 919.3 0.18
3 2.201 5.7 26.76 1060 1033.24 0.2
4 1.748 7.47 39.92 1256.27 1216.35 39.58
main
w1
w2
w3
w4
    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.