T3 Code · Performance over time

Main thread user CPU

All metrics
21 / 120
Main thread user CPU (lower is better)
25.9% of one core
18 Nov−0.6% vs the median of up to 7 previous complete nightlies, within noiseFork perf.9 18.1%, −33.1% 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.

User CPU time of the server's main thread over the window.

Lower is better. Noise band 24.6% on Linux x64 (n = 6), so a change counts beyond 29.6%.

Unit
% of one core
Better
lower
Noise band
24.6%, threshold 29.6% (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
Every upstream build, Linux only
Source
result.json: metrics.cpu.main.user_pct

Main thread user CPU (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.20261118.490325.9%25.7 to 26.13−0.6%within noise
17 Nov 2026v0.1.11-nightly.20261117.485126.1%26.0 to 27.03+4.3%within noise
16 Nov 2026v0.1.10-nightly.20261116.480026.5%25.9 to 27.43+22.7%regressed beyond noise
15 Nov 2026v0.1.10-nightly.20261115.474926.9%26.2 to 26.93+48.3%regressed beyond noise
14 Nov 2026v0.1.10-nightly.20261114.470927.5%25.9 to 27.63+52.7%regressed beyond noise
13 Nov 2026v0.1.10-nightly.20261113.465625.0%24.4 to 25.33+40.3%regressed beyond noise
12 Nov 2026v0.1.10-nightly.20261112.460622.0%21.8 to 23.43+23.3%regressed beyond noise
11 Nov 2026v0.1.10-nightly.20261111.454419.7%18.1 to 21.03+10.4%regressed beyond noise
10 Nov 2026v0.1.9-nightly.20261110.449119.8%19.4 to 21.23+10.9%regressed beyond noise
9 Nov 2026v0.1.9-nightly.20261109.444317.7%17.2 to 18.53−0.8%within noise
8 Nov 2026v0.1.8-nightly.20261108.438017.0%16.4 to 17.73−4.8%within noise
7 Nov 2026v0.1.8-nightly.20261107.434017.3%16.9 to 17.43−3.1%within noise
6 Nov 2026v0.1.8-nightly.20261106.428817.0%16.7 to 18.13−4.8%within noise
5 Nov 2026v0.1.8-nightly.20261105.423417.8%17.6 to 18.33−2.2%within noise, thin reference
4 Nov 2026v0.1.8-nightly.20261104.419218.2%18.1 to 18.23+1.9%within noise, thin reference
3 Nov 2026v0.1.7-nightly.20261103.413817.8%17.5 to 18.03no reference yet
2 Nov 2026v0.1.7-nightly.20261102.409318.6%18.3 to 19.03no reference yet
1 Nov 2026v0.1.7-nightly.20261101.404817.4%16.9 to 18.73no reference yet
31 Oct 2026v0.1.7-nightly.20261031.400418.0%18.0 to 18.13no reference yet
30 Oct 2026v0.1.6-nightly.20261030.394017.6%17.2 to 17.83no reference yet
29 Oct 2026v0.1.6-nightly.20261029.389215.5%15.2 to 15.53no reference yet
28 Oct 2026v0.1.4-nightly.20261028.384115.4%15.2 to 15.73no reference yet
27 Oct 2026v0.1.4-nightly.20261027.379616.5%16.1 to 16.63no reference yet
26 Oct 2026v0.1.3-nightly.20261026.374016.8%16.1 to 16.93no reference yet
25 Oct 2026v0.1.3-nightly.20261025.368214.6%14.2 to 14.83no reference yet
24 Oct 2026v0.1.3-nightly.20261024.364314.9%14.9 to 15.03no reference yet
23 Oct 2026v0.1.3-nightly.20261023.359215.6%15.5 to 17.33no reference yet
22 Oct 2026v0.1.2-nightly.20261022.354121%21.0 to 21.23no reference yet
21 Oct 2026v0.1.2-nightly.20261021.349918.8%18.7 to 19.03no reference yet
20 Oct 2026v0.1.1-nightly.20261020.343818.9%18.8 to 19.33no reference yet
19 Oct 2026v0.1.1-nightly.20261019.337326.2%25.1 to 29.53no noise band yet, no verdict
18 Oct 2026v0.1.1-nightly.20261018.333525.3%24.4 to 26.13no noise band yet, no verdict
17 Oct 2026v0.1.1-nightly.20261017.327224.3%24.3 to 24.73no noise band yet, no verdict
Ladder as a table
AgentsNightly 18 NovFork perf.9
13.10%2.85%
5025.9%18.1%
10019.0%30.3%
15022.2%39.1%
20026.9%47.4%
25025.9%58.1%
Download this series