Skip to content

Conversation

ErichDonGubler
Copy link
Contributor

No description provided.

Copy link

github-actions bot commented May 24, 2025

Failed Interop Tests

QUIC Interop Runner, client vs. server, differences relative to 5e93294.

neqo-latest as client

neqo-latest as server

All results

Succeeded Interop Tests

QUIC Interop Runner, client vs. server

neqo-latest as client

neqo-latest as server

Unsupported Interop Tests

QUIC Interop Runner, client vs. server

neqo-latest as client

neqo-latest as server

@ErichDonGubler
Copy link
Contributor Author

The last commit might need to be omitted, pending proof that the latest version actually compiles.

Copy link

Benchmark results

1-conn/1-100mb-resp/mtu-1504 (aka. Download)/client: Change within noise threshold.
       time:   [760.94 ms 788.79 ms 818.47 ms]
       thrpt:  [122.18 MiB/s 126.78 MiB/s 131.42 MiB/s]
change:
       time:   [−11.209% −6.0606% −0.6973%] (p = 0.03 < 0.05)
       thrpt:  [+0.7022% +6.4516% +12.624%]

Found 22 outliers among 100 measurements (22.00%)
22 (22.00%) high severe

1-conn/10_000-parallel-1b-resp/mtu-1504 (aka. RPS)/client: No change in performance detected.
       time:   [314.65 ms 315.96 ms 317.27 ms]
       thrpt:  [31.519 Kelem/s 31.649 Kelem/s 31.781 Kelem/s]
change:
       time:   [−0.9201% −0.3799% +0.1884%] (p = 0.18 > 0.05)
       thrpt:  [−0.1881% +0.3813% +0.9286%]

Found 5 outliers among 100 measurements (5.00%)
3 (3.00%) low mild
2 (2.00%) high mild

1-conn/1-1b-resp/mtu-1504 (aka. HPS)/client: No change in performance detected.
       time:   [34.545 ms 35.787 ms 37.027 ms]
       thrpt:  [27.007  elem/s 27.943  elem/s 28.948  elem/s]
change:
       time:   [−4.8421% +0.0685% +5.2570%] (p = 0.98 > 0.05)
       thrpt:  [−4.9945% −0.0685% +5.0885%]
1-conn/1-100mb-req/mtu-1504 (aka. Upload)/client: Change within noise threshold.
       time:   [7.3247 s 7.3374 s 7.3500 s]
       thrpt:  [13.605 MiB/s 13.629 MiB/s 13.652 MiB/s]
change:
       time:   [+0.5088% +0.7758% +1.0342%] (p = 0.00 < 0.05)
       thrpt:  [−1.0236% −0.7699% −0.5062%]
decode 4096 bytes, mask ff: 💚 Performance has improved.
       time:   [11.636 µs 11.648 µs 11.660 µs]
       change: [−1.8845% −1.7549% −1.6211%] (p = 0.00 < 0.05)

Found 2 outliers among 100 measurements (2.00%)
2 (2.00%) high mild

decode 1048576 bytes, mask ff: 💚 Performance has improved.
       time:   [2.9767 ms 2.9790 ms 2.9811 ms]
       change: [−1.9004% −1.7426% −1.5899%] (p = 0.00 < 0.05)

Found 2 outliers among 100 measurements (2.00%)
1 (1.00%) high mild
1 (1.00%) high severe

decode 4096 bytes, mask 7f: 💚 Performance has improved.
       time:   [19.425 µs 19.438 µs 19.450 µs]
       change: [−2.9813% −2.8102% −2.6426%] (p = 0.00 < 0.05)

Found 3 outliers among 100 measurements (3.00%)
1 (1.00%) high mild
2 (2.00%) high severe

decode 1048576 bytes, mask 7f: 💚 Performance has improved.
       time:   [4.9749 ms 4.9770 ms 4.9788 ms]
       change: [−2.6305% −2.4697% −2.2811%] (p = 0.00 < 0.05)

Found 14 outliers among 100 measurements (14.00%)
4 (4.00%) low mild
10 (10.00%) high severe

decode 4096 bytes, mask 3f: 💚 Performance has improved.
       time:   [5.5515 µs 5.5592 µs 5.5668 µs]
       change: [−33.227% −33.106% −32.981%] (p = 0.00 < 0.05)

Found 2 outliers among 100 measurements (2.00%)
2 (2.00%) high mild

decode 1048576 bytes, mask 3f: 💚 Performance has improved.
       time:   [1.4232 ms 1.4253 ms 1.4275 ms]
       change: [−30.959% −29.541% −28.037%] (p = 0.00 < 0.05)

Found 12 outliers among 100 measurements (12.00%)
2 (2.00%) high mild
10 (10.00%) high severe

1000 streams of 1 bytes/multistream: 💚 Performance has improved.
       time:   [14.107 ns 14.464 ns 15.145 ns]
       change: [−15.612% −13.389% −9.3107%] (p = 0.00 < 0.05)

Found 9 outliers among 500 measurements (1.80%)
3 (0.60%) low mild
4 (0.80%) high mild
2 (0.40%) high severe

1000 streams of 1000 bytes/multistream: 💚 Performance has improved.
       time:   [14.015 ns 14.053 ns 14.093 ns]
       change: [−13.837% −13.450% −13.061%] (p = 0.00 < 0.05)

Found 4 outliers among 500 measurements (0.80%)
3 (0.60%) high mild
1 (0.20%) high severe

coalesce_acked_from_zero 1+1 entries: No change in performance detected.
       time:   [95.021 ns 95.180 ns 95.330 ns]
       change: [−0.6872% −0.0572% +0.5809%] (p = 0.87 > 0.05)

Found 10 outliers among 100 measurements (10.00%)
5 (5.00%) low severe
5 (5.00%) low mild

coalesce_acked_from_zero 3+1 entries: No change in performance detected.
       time:   [112.86 ns 112.99 ns 113.10 ns]
       change: [−0.2025% +0.3973% +0.9397%] (p = 0.16 > 0.05)

Found 10 outliers among 100 measurements (10.00%)
10 (10.00%) low mild

coalesce_acked_from_zero 10+1 entries: No change in performance detected.
       time:   [111.62 ns 111.81 ns 111.98 ns]
       change: [−0.4717% +0.1017% +0.6741%] (p = 0.74 > 0.05)
coalesce_acked_from_zero 1000+1 entries: No change in performance detected.
       time:   [89.114 ns 89.544 ns 90.407 ns]
       change: [−2.1253% −0.4974% +0.5219%] (p = 0.66 > 0.05)

Found 4 outliers among 100 measurements (4.00%)
2 (2.00%) high mild
2 (2.00%) high severe

RxStreamOrderer::inbound_frame(): Change within noise threshold.
       time:   [111.26 ms 111.27 ms 111.29 ms]
       change: [+0.4052% +0.4261% +0.4474%] (p = 0.00 < 0.05)

Found 3 outliers among 100 measurements (3.00%)
2 (2.00%) low mild
1 (1.00%) high mild

SentPackets::take_ranges: No change in performance detected.
       time:   [9.4279 µs 9.4897 µs 9.5519 µs]
       change: [−3.9277% −1.2850% +1.4468%] (p = 0.35 > 0.05)

Found 4 outliers among 100 measurements (4.00%)
4 (4.00%) low severe

transfer/pacing-false/varying-seeds: Change within noise threshold.
       time:   [34.043 ms 34.074 ms 34.106 ms]
       change: [+1.1928% +1.3325% +1.4701%] (p = 0.00 < 0.05)
transfer/pacing-true/varying-seeds: No change in performance detected.
       time:   [34.698 ms 34.746 ms 34.796 ms]
       change: [−0.0978% +0.0951% +0.2971%] (p = 0.36 > 0.05)

Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) high mild

transfer/pacing-false/same-seed: Change within noise threshold.
       time:   [34.104 ms 34.125 ms 34.147 ms]
       change: [+1.1873% +1.2965% +1.4053%] (p = 0.00 < 0.05)
transfer/pacing-true/same-seed: Change within noise threshold.
       time:   [36.004 ms 36.032 ms 36.059 ms]
       change: [+1.9469% +2.0751% +2.2024%] (p = 0.00 < 0.05)

Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) low mild

Client/server transfer results

Transfer of 33554432 bytes over loopback, min. 100 runs. All unit-less numbers are in milliseconds.

Client vs. server (params) Mean ± σ Min Max MiB/s ± σ Δ main Δ main
google vs. google 535.6 ± 65.2 495.4 906.8 59.7 ± 0.5
google vs. neqo (cubic, paced) 390.6 ± 48.5 356.1 689.3 81.9 ± 0.7
msquic vs. msquic 174.6 ± 55.2 135.3 370.4 183.3 ± 0.6
msquic vs. neqo (cubic, paced) 293.1 ± 48.8 266.3 528.1 109.2 ± 0.7
neqo vs. google (cubic, paced) 823.8 ± 43.1 782.5 1034.9 38.8 ± 0.7
neqo vs. msquic (cubic, paced) 219.0 ± 48.6 191.6 452.2 146.1 ± 0.7
neqo vs. neqo (cubic) 265.9 ± 59.1 234.8 650.2 120.4 ± 0.5
neqo vs. neqo (cubic, paced) 266.8 ± 53.8 233.6 491.8 119.9 ± 0.6
neqo vs. neqo (reno) 278.0 ± 72.7 217.7 737.0 115.1 ± 0.4
neqo vs. neqo (reno, paced) 266.8 ± 55.9 235.7 534.7 119.9 ± 0.6
neqo vs. quiche (cubic, paced) 261.2 ± 51.5 230.5 479.9 122.5 ± 0.6
neqo vs. s2n (cubic, paced) 272.6 ± 53.0 244.3 490.4 117.4 ± 0.6
quiche vs. neqo (cubic, paced) 419.2 ± 48.5 379.8 655.4 76.3 ± 0.7
quiche vs. quiche 221.7 ± 68.5 173.9 580.5 144.4 ± 0.5
s2n vs. neqo (cubic, paced) 318.6 ± 45.5 285.5 542.3 100.5 ± 0.7
s2n vs. s2n 273.1 ± 68.8 236.6 569.5 117.2 ± 0.5

⬇️ Download logs

Copy link
Member

@mxinden mxinden left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you for doing this work across so many projects. I am fine with Neqo supporting a windows version range.

@ErichDonGubler
Copy link
Contributor Author

Just getting back from PTO! Will tackle this now. 🙂

@ErichDonGubler ErichDonGubler force-pushed the windows-0.59 branch 2 times, most recently from c2bd68a to 6e8e538 Compare June 6, 2025 06:26
@mxinden
Copy link
Member

mxinden commented Jun 10, 2025

Let us know once you need another review :)

@larseggert
Copy link
Collaborator

@ErichDonGubler is this reviewable yet?

@ErichDonGubler
Copy link
Contributor Author

ErichDonGubler commented Jul 3, 2025

@larseggert: Not yet! mio and libloading have made releases allowing Windows 0.59, while tokio and backtrace are still lagging. See also the index comment I'm keeping up-to-date at gfx-rs/wgpu#6876 (comment).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants