T3 Code · Performance over time

Main thread busy

All metrics
20 / 120
Main thread busy (lower is better)
68.4% of one core
18 Nov−0.6% vs the median of up to 7 previous complete nightlies, within noiseFork perf.9 19.6%, −71.6% 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 plus system CPU time of the server's main thread over the window, as a share of one core.

Lower is better. Noise band 22.3% on Linux x64 (n = 6), so a change counts beyond 27.3%.

Unit
% of one core
Better
lower
Noise band
22.3%, threshold 27.3% (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.busy_pct

Main thread busy (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.20261118.490368.4%67.4 to 68.73−0.6%within noise
17 Nov 2026v0.1.11-nightly.20261117.485170.2%69.2 to 71.23+2.8%within noise
16 Nov 2026v0.1.10-nightly.20261116.480068.3%64.9 to 71.83+34.3%regressed beyond noise
15 Nov 2026v0.1.10-nightly.20261115.474968.8%67.6 to 69.03+108.5%regressed beyond noise
14 Nov 2026v0.1.10-nightly.20261114.470969.8%69.4 to 70.33+113.3%regressed beyond noise
13 Nov 2026v0.1.10-nightly.20261113.465669.2%68.7 to 70.33+112.9%regressed beyond noise
12 Nov 2026v0.1.10-nightly.20261112.460659.6%56.7 to 60.43+83.2%regressed beyond noise
11 Nov 2026v0.1.10-nightly.20261111.454436.1%32.5 to 37.33+10.9%regressed beyond noise
10 Nov 2026v0.1.9-nightly.20261110.449135.8%35.0 to 38.43+10.1%regressed beyond noise
9 Nov 2026v0.1.9-nightly.20261109.444332.4%32.2 to 34.13−0.3%within noise
8 Nov 2026v0.1.8-nightly.20261108.438032.6%29.7 to 33.33+0.2%within noise
7 Nov 2026v0.1.8-nightly.20261107.434031.8%31.7 to 31.83−2.0%within noise
6 Nov 2026v0.1.8-nightly.20261106.428832.3%32.2 to 34.53−0.5%within noise
5 Nov 2026v0.1.8-nightly.20261105.423431.4%31.3 to 33.43−4.7%within noise, thin reference
4 Nov 2026v0.1.8-nightly.20261104.419233.0%32.4 to 33.13+1.4%within noise, thin reference
3 Nov 2026v0.1.7-nightly.20261103.413832.5%31.9 to 32.73no reference yet
2 Nov 2026v0.1.7-nightly.20261102.409333.4%32.7 to 34.03no reference yet
1 Nov 2026v0.1.7-nightly.20261101.404832.3%32.1 to 35.13no reference yet
31 Oct 2026v0.1.7-nightly.20261031.400432.6%32.0 to 32.93no reference yet
30 Oct 2026v0.1.6-nightly.20261030.394033.0%31.7 to 33.63no reference yet
29 Oct 2026v0.1.6-nightly.20261029.389227.2%26.8 to 27.73no reference yet
28 Oct 2026v0.1.4-nightly.20261028.384127.7%27.5 to 27.83no reference yet
27 Oct 2026v0.1.4-nightly.20261027.379629.0%28.4 to 29.03no reference yet
26 Oct 2026v0.1.3-nightly.20261026.374030.6%28.8 to 30.73no reference yet
25 Oct 2026v0.1.3-nightly.20261025.368227.0%26.7 to 27.73no reference yet
24 Oct 2026v0.1.3-nightly.20261024.364327.4%26.7 to 28.13no reference yet
23 Oct 2026v0.1.3-nightly.20261023.359228.0%27.9 to 31.43no reference yet
22 Oct 2026v0.1.2-nightly.20261022.354134.9%33.8 to 35.03no reference yet
21 Oct 2026v0.1.2-nightly.20261021.349932.6%32.4 to 33.13no reference yet
20 Oct 2026v0.1.1-nightly.20261020.343832.9%32.9 to 33.53no reference yet
19 Oct 2026v0.1.1-nightly.20261019.337348.3%46.2 to 51.83no noise band yet, no verdict
18 Oct 2026v0.1.1-nightly.20261018.333546.8%45.0 to 48.03no noise band yet, no verdict
17 Oct 2026v0.1.1-nightly.20261017.327245.7%45.4 to 45.83no noise band yet, no verdict
Ladder as a table
AgentsNightly 18 NovFork perf.9
14.78%3.16%
5068.4%19.6%
10069.2%32.7%
15072.4%42.0%
20074.0%51.0%
25073.3%62.3%
Download this series