This report contains the final payload and batch measurements: UDP on 2026-10-12, and other protocols and matched native TCP executions on 2026-10-10. Each protocol retains its own source hash. Coroutine results measure native TCP.
Local Release IPv4 loopback: 16 clients, three payload sizes, four windows and three thread models, with three repetitions per configuration.
| Configuration | Test machine |
|---|---|
| Machine model | Mac16,11 |
| Processor | Apple M4 Pro |
| Architecture | arm64 |
| Physical / logical CPUs | 14 / 14 |
| Memory | 48 GiB |
| Operating system | macOS 27.0.1 (26A434) |
| Kernel release | 27.0.0 |
Measurement 1.0 s, warmup 0.25 s. Charts show medians and min/max, not confidence intervals; p99 is the median of per-run sampled percentiles.
Average process CPU uses 100% for one logical CPU and can exceed 100% with multiple threads; estimated values are labeled, and incomplete sets of three CPU samples show N/A. On this 14-logical-CPU host, machine share = process CPU% / 14; 100% of one CPU is 7.14% of the machine.
- Source SHA-256:
fa69d6ce97745281290a263cb0f2515470ede0edb77f5132ce6e3ac7ae7cd9b4 - UDP source SHA-256:
0771574af8ca56f1914deb544b956079e0a2f24021953ded5f5d524e3df334b1; raw records and conditions. - Measurements: 2,376; valid profiles: 782; failed profiles: 10.
Apple Clang 21.0.0; CMake 4.3.3; Ninja; Release -O3 -DNDEBUG -std=gnu++20 -arch arm64; TLS ON; sanitizers OFF. Standalone Asio 1.38.2 with BHO Beast based on Boost 1.84 (API 351); Boost 1.90.0 with Asio 1.38.0 and Beast API 359; OpenSSL 3.6.4. Mac mini on AC power, low power mode disabled, no CPU affinity.
The arknet protocol program measures message RTT; matched native Asio callback/coroutine executions measure batch RTT. Their fixed-batch results do not establish speedups over arknet endpoints; public coroutine endpoint APIs remain TODO.
TCP
Charts, findings and failed profiles
UDP
Charts, findings and failed profiles
WebSocket
Charts, findings and failed profiles
TCPS
Charts, findings and failed profiles
WSS
Charts, findings and failed profiles
HTTP
Charts, findings and failed profiles
HTTPS
Charts, findings and failed profiles
Native TCP
Charts, findings and failed profiles
Raw Data
- standalone-callback.json: 756 measurements; executable SHA-256
480122707b4b3ed592d1f580f5cf7092b8008ce52331f55f0e666e89c568be6b - boost-callback.json: 756 measurements; executable SHA-256
7b2989fd23e15a68a662899d1906f3b094138018b8ada5904309b4a90d96ea34 - standalone-native.json: 432 measurements; executable SHA-256
5d7ac94153f942eda1494656edc865ab01817a7ea9122a240cc5f3260b061549 - boost-native.json: 432 measurements; executable SHA-256
d68dc3aea70b8334e5fc88d3bda0739faeb8696842c4f2f582092d92cb1b2d6d
Reproduction and metric definitions
Local loopback, fixed connection distributions and short runs do not establish production capacity; repeat with real messages and connection activity.