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
1235 ms
total CPU used
5314.79 ms
speedup
4.3×
vs serial
efficiency
86%
of 5× ideal
stream spawn ms spawned@ work start@ work end@ work ms reap wait ms
0 (main) 0 7.2 7.21 1079.23 1072.02 0
1 2.322 2.33 14.84 1124.26 1109.42 45.16
2 1.604 3.96 16.51 982.9 966.39 0.11
3 1.617 5.59 23.63 1231.94 1208.31 152.88
4 1.58 7.19 38.11 996.76 958.65 0.17
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.