CPU scaling benchmark
workers
3 +1 main
iters total
500M
125000000/stream
elapsed
1117.27 ms
total CPU used
4356.49 ms
speedup
3.9×
vs serial
efficiency
97.5%
of 4× ideal
| stream | spawn ms | spawned@ | work start@ | work end@ | work ms | reap wait ms |
|---|---|---|---|---|---|---|
| 0 (main) | 0 | 5.72 | 5.72 | 1086.86 | 1081.14 | 0 |
| 1 | 2.09 | 2.11 | 14.48 | 1104.95 | 1090.47 | 18.25 |
| 2 | 1.567 | 3.69 | 15.61 | 1106.01 | 1090.4 | 21.14 |
| 3 | 2.002 | 5.71 | 18.45 | 1112.93 | 1094.48 | 26.17 |
main
w1
w2
w3
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.