Less memory under load.
727.78 vs. 4,102.04 MiB process-tree RSS with 50 active static Pages.
Memory under load, processing speed at equal Page concurrency and network acquisition with Optimize. Controlled workloads with explicit correctness checks.
727.78 vs. 4,102.04 MiB process-tree RSS with 50 active static Pages.
164.62 vs. 22.11 Pages/s at the same 50-Page concurrency, including create and teardown.
284,591 → 5,276 HTTP body bytes for Books extraction with Optimize on versus off. Separate workload; 5/5 passes.
250/250 measured static batch attempts passed for each runtime. These runs are separate observations, with a small instrumentation-boundary difference recorded below.
CPU cost: 69.94 vs. 133.38 ms per Page. Mimic batch latency p50/p95: 161.33 / 197.39 ms; Chrome: 1,853.23 / 2,382.80 ms. Batch speed does not imply faster single-Page completion.
Measured active RSS, batch throughput, CPU per Page and RSS 250 ms after teardown. A failed series is excluded from successful comparisons. An absent Chrome value means no successful matching reference is available.
| Workload | Pages | Mimic / Chrome RSS MiB | Mimic / Chrome Pages/s | Mimic / Chrome CPU ms/Page | Mimic recovery RSS MiB | Valid / attempts |
|---|---|---|---|---|---|---|
| static | 1 | 106.44 / 1210.81 | 16.87 / 8.63 | 38.28 / 222.66 | 88.05 | 20 / 20 |
| static | 5 | 148.94 / 1373.74 | 80.55 / 22.15 | 51.88 / 130.63 | 85.48 | 25 / 25 |
| static | 10 | 213.13 / 1682.53 | 79.79 / 24.19 | 61.88 / 126.56 | 101.89 | 50 / 50 |
| static | 25 | 418.23 / 2571.97 | 164.27 / 25.03 | 59.13 / 123.25 | 117.70 | 125 / 125 |
| static | 50 | 727.78 / 4102.04 | 164.62 / 22.11 | 69.94 / 133.38 | 133.59 | 250 / 250 |
| static | 100 | 1351.21 / — | 170.15 / 21.67 | 75.63 / 136.41 | 169.20 | 500 / 500 |
| cpu | 1 | 124.77 / — | 9.19 / 8.66 | 81.25 / 234.38 | 96.77 | 20 / 20 |
| cpu | 5 | 206.46 / — | 43.22 / 20.89 | 87.50 / 161.88 | 92.16 | 25 / 25 |
| cpu | 10 | 316.48 / — | 74.82 / 26.85 | 97.19 / 170.31 | 112.44 | 50 / 50 |
| cpu | 25 | 671.74 / — | 95.82 / 29.18 | 142.88 / 159.50 | 122.14 | 125 / 125 |
| cpu | 50 | 1225.69 / — | 111.25 / 29.50 | 147.38 / 154.50 | 141.21 | 250 / 250 |
| cpu | 100 | 2333.55 / — | 108.20 / 19.14 | 157.28 / 194.75 | 184.64 | 500 / 500 |
| react | 1 | 119.95 / — | 9.22 / 6.87 | 103.91 / 319.53 | 97.69 | 20 / 20 |
| react | 5 | 180.53 / — | 43.64 / 5.89 | 121.25 / 230.63 | 94.63 | 25 / 25 |
| react | 10 | 288.28 / — | 56.41 / 8.65 | 143.13 / 186.88 | 123.61 | 50 / 50 |
| react | 25 | 569.34 / — | 75.96 / 24.76 | 174.50 / 157.25 | 159.75 | 125 / 125 |
| react | 50 | 1013.87 / — | 83.26 / 17.08 | 180.81 / 194.63 | 208.83 | 250 / 250 |
| react | 100 | — / — | — / 16.57 | — / 212.44 | — | 400 / 500 · failed |
RSS sums process working sets and can count shared pages more than once. Recovery is a short observation, not proof of leak absence or an absolute capacity limit.
20 measured warm attempts per workload and runtime, one warmup excluded. These tables include workloads where Mimic completes more slowly. CPU covers Page creation through teardown; completion spans navigation through the exact workload result.
| Workload | Mimic p50 / p95 ms | Chrome p50 / p95 ms | Mimic / Chrome CPU ms/Page |
|---|---|---|---|
| static | 32.05 / 35.02 | 18.44 / 22.07 | 31.25 / 140.63 |
| cpu | 59.47 / 65.83 | 45.59 / 52.14 | 62.50 / 234.38 |
| dom | 155.73 / 163.30 | 48.28 / 66.57 | 187.50 / 210.94 |
| async | 71.39 / 82.74 | 50.98 / 55.15 | 85.94 / 203.13 |
| react | 66.33 / 69.23 | 44.95 / 51.90 | 93.75 / 250.00 |
| wasm | 30.96 / 37.61 | 23.13 / 30.59 | 46.88 / 179.69 |
Books SSR extraction: 98.1% fewer acquired HTTP body bytes, 284,591 → 5,276; transport responses 29 → 1. Both Mimic Default and installed Auto passed 5/5 matched local trials. Manual tied Auto. Training and validation took 27.1 seconds.
Separate recorded workload from October 5. Encoded acquired bodies are not physical wire traffic. One captured state does not establish savings on other sites or future states.
Optimize results and limitsThe final Mimic executable passed 120/120 warm single-Page attempts. Static and CPU completed 100 Pages; React completed 50 Pages. React-100 failed in its final wave with a native access violation and is excluded from successful comparisons: 2,810/2,910 density attempts valid.
Active-memory comparisons reuse the frozen Chrome reference; timing/CPU use the saved October 10 Chrome capture. Both runners use the same fixture, frozen execute, navigation barrier and 50 ms completion polling. Candidate wave timing includes the initial accounting snapshot and excludes the final one; reference timing does the reverse. Guards differ (120 vs. 180 seconds) and were not reached. Runs are unpaired; background load can vary. CPU is process-tree user+kernel time; latency quantiles use linear interpolation.
Mimic does not render pixels. These fixtures do not establish general website compatibility or a universal speedup. Exact binary, fixture, adapter and capture hashes are available in the methodology and data.