T3 Code · Performance over time

Seconds with a stall of 40 ms or more

All metrics
34 / 120
Seconds with a stall of 40 ms or more (lower is better)
2.01%
18 Nov+19.6% vs the median of up to 7 previous complete nightlies, within noiseFork perf.9 0%, −100.0% vs its upstream base
AgentsRange

When upstream merges a fix, the lines meet.

Upstream nightlyPerformance forkNoise bandRange of attempts green = better, red = worse, beyond the noise band

As agents grow.

Share of the window's 1 s intervals whose event loop delay maximum reached 40 ms.

Lower is better. Noise band 140.4% on Linux x64 (n = 6), so a change counts beyond 145.4%.

Unit
%
Better
lower
Noise band
140.4%, threshold 145.4% (n = 6)
Recorded from
17 Oct 2026; before that the series is a gap, never a zero
Platforms
Linux x64
Lane
Every record, inside the agent ladder
Upstream builds
Needs the harness's preload
Source
loop-1hz.ndjson: share of window lines with delay.max_ms >= 40

Seconds with a stall of 40 ms or more (lower is better) at 1 to 250 agents, Linux x64

against
Records, 50 agents
DateTagMedianMin to maxValidvs up to 7 previousVerdict
18 Nov 2026v0.1.11-nightly.20261118.49032.01%1.00 to 2.013+19.6%within noise
17 Nov 2026v0.1.11-nightly.20261117.48511.67%1.67 to 2.343−16.9%within noise
16 Nov 2026v0.1.10-nightly.20261116.48001.68%1.67 to 2.343−44.2%within noise
15 Nov 2026v0.1.10-nightly.20261115.47491.67%1.67 to 2.013−58.2%within noise
14 Nov 2026v0.1.10-nightly.20261114.47091.34%0.67 to 1.343−73.3%within noise
13 Nov 2026v0.1.10-nightly.20261113.46562.01%1.34 to 2.013−66.6%within noise
12 Nov 2026v0.1.10-nightly.20261112.460685.5%84.3 to 923+1320.4%regressed beyond noise
11 Nov 2026v0.1.10-nightly.20261111.454413.4%9.03 to 15.73+122.2%regressed beyond noise
10 Nov 2026v0.1.9-nightly.20261110.449116.3%13.7 to 16.83+171.3%regressed beyond noise
9 Nov 2026v0.1.9-nightly.20261109.44438.36%7.02 to 15.73+38.9%within noise
8 Nov 2026v0.1.8-nightly.20261108.43809.70%3 to 11.73+61.1%within noise
7 Nov 2026v0.1.8-nightly.20261107.43405.69%5 to 6.673−5.5%within noise
6 Nov 2026v0.1.8-nightly.20261106.42886.69%6 to 11.43+11.1%within noise
5 Nov 2026v0.1.8-nightly.20261105.42344%3.33 to 6.363−37.1%within noise, thin reference
4 Nov 2026v0.1.8-nightly.20261104.41926.36%5 to 8.033+5.6%within noise, thin reference
3 Nov 2026v0.1.7-nightly.20261103.41386.02%5 to 8.033no reference yet
2 Nov 2026v0.1.7-nightly.20261102.40935.35%4.01 to 6.023no reference yet
1 Nov 2026v0.1.7-nightly.20261101.40486.02%5 to 10.33no reference yet
31 Oct 2026v0.1.7-nightly.20261031.40046.02%5.67 to 6.693no reference yet
30 Oct 2026v0.1.6-nightly.20261030.394010.4%7.33 to 13.03no reference yet
29 Oct 2026v0.1.6-nightly.20261029.38925.67%5.33 to 63no reference yet
28 Oct 2026v0.1.4-nightly.20261028.38415.67%5.67 to 63no reference yet
27 Oct 2026v0.1.4-nightly.20261027.37966%5.69 to 6.023no reference yet
26 Oct 2026v0.1.3-nightly.20261026.37406.02%5.67 to 6.023no reference yet
25 Oct 2026v0.1.3-nightly.20261025.36825.33%4.67 to 5.353no reference yet
24 Oct 2026v0.1.3-nightly.20261024.36435.02%5 to 63no reference yet
23 Oct 2026v0.1.3-nightly.20261023.35921.67%1.34 to 3.013no reference yet
22 Oct 2026v0.1.2-nightly.20261022.354130.8%29.4 to 33.43no reference yet
21 Oct 2026v0.1.2-nightly.20261021.349922.7%22.7 to 253no reference yet
20 Oct 2026v0.1.1-nightly.20261020.343822.4%22.1 to 25.83no reference yet
19 Oct 2026v0.1.1-nightly.20261019.337344.0%37.9 to 51.03no noise band yet, no verdict
18 Oct 2026v0.1.1-nightly.20261018.333539.8%34.6 to 40.53no noise band yet, no verdict
17 Oct 2026v0.1.1-nightly.20261017.327237.1%36.2 to 38.33no noise band yet, no verdict
Ladder as a table
AgentsNightly 18 NovFork perf.9
10%0%
502.01%0%
1003.69%2.33%
1502.68%2%
2004.36%2%
2505.03%3.33%
Download this series