T3 Code · Performance over time

Worker threads CPU

All metrics
24 / 120
Worker threads CPU (lower is better)
27.4% of one core
18 Nov−0.5% vs the median of up to 7 previous complete nightlies, within noiseFork perf.9 9.56%, −66.0% 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.

CPU time of the server's worker threads (SQLite readers and the checkpoint worker).

Lower is better. Noise band 25.2% on Linux x64 (n = 6), so a change counts beyond 30.2%.

Unit
% of one core
Better
lower
Noise band
25.2%, threshold 30.2% (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.worker_threads.busy_pct

Worker threads 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.20261118.490327.4%27.1 to 27.63−0.5%within noise
17 Nov 2026v0.1.11-nightly.20261117.485128.2%28.0 to 28.43+5.6%within noise
16 Nov 2026v0.1.10-nightly.20261116.480026.7%26.0 to 28.33+30.9%regressed beyond noise
15 Nov 2026v0.1.10-nightly.20261115.474928.1%27.3 to 28.23+113.3%regressed beyond noise
14 Nov 2026v0.1.10-nightly.20261114.470928.1%27 to 28.73+117.5%regressed beyond noise
13 Nov 2026v0.1.10-nightly.20261113.465627.6%27.1 to 28.23+117.2%regressed beyond noise
12 Nov 2026v0.1.10-nightly.20261112.460613.1%12.8 to 14.33+3.3%within noise
11 Nov 2026v0.1.10-nightly.20261111.454413.5%12.4 to 14.53+6.5%within noise
10 Nov 2026v0.1.9-nightly.20261110.449114.0%13.4 to 15.43+10.0%within noise
9 Nov 2026v0.1.9-nightly.20261109.444312.9%12.3 to 14.13+1.2%within noise
8 Nov 2026v0.1.8-nightly.20261108.438012.3%11.5 to 12.53−2.9%within noise
7 Nov 2026v0.1.8-nightly.20261107.434012.5%12.2 to 12.63−1.5%within noise
6 Nov 2026v0.1.8-nightly.20261106.428812.5%12.3 to 12.73−1.2%within noise
5 Nov 2026v0.1.8-nightly.20261105.423412.7%12.4 to 12.73−0.3%within noise, thin reference
4 Nov 2026v0.1.8-nightly.20261104.419212.7%12.7 to 12.93−3.7%within noise, thin reference
3 Nov 2026v0.1.7-nightly.20261103.413813.2%12.7 to 13.33no reference yet
2 Nov 2026v0.1.7-nightly.20261102.409317.2%16.9 to 17.23no reference yet
1 Nov 2026v0.1.7-nightly.20261101.404820.7%20.7 to 21.13no reference yet
31 Oct 2026v0.1.7-nightly.20261031.400420.4%20.1 to 20.63no reference yet
30 Oct 2026v0.1.6-nightly.20261030.394020.8%20.7 to 21.23no reference yet
29 Oct 2026v0.1.6-nightly.20261029.389214.6%14.3 to 14.93no reference yet
28 Oct 2026v0.1.4-nightly.20261028.384114.7%14.5 to 14.73no reference yet
27 Oct 2026v0.1.4-nightly.20261027.379615.1%14.9 to 15.33no reference yet
26 Oct 2026v0.1.3-nightly.20261026.374015.1%15.0 to 15.33no reference yet
25 Oct 2026v0.1.3-nightly.20261025.368213.6%13.0 to 13.83no reference yet
24 Oct 2026v0.1.3-nightly.20261024.364313.7%13.5 to 13.73no reference yet
23 Oct 2026v0.1.3-nightly.20261023.359213.8%13.7 to 14.53no reference yet
22 Oct 2026v0.1.2-nightly.20261022.35410.48%0.47 to 0.493no reference yet
21 Oct 2026v0.1.2-nightly.20261021.34990.50%0.49 to 0.523no reference yet
20 Oct 2026v0.1.1-nightly.20261020.34380.50%0.50 to 0.513no reference yet
19 Oct 2026v0.1.1-nightly.20261019.33731.33%1.23 to 1.343no noise band yet, no verdict
18 Oct 2026v0.1.1-nightly.20261018.33351.23%1.23 to 1.303no noise band yet, no verdict
17 Oct 2026v0.1.1-nightly.20261017.32721.24%1.23 to 1.293no noise band yet, no verdict
Ladder as a table
AgentsNightly 18 NovFork perf.9
11.87%1.36%
5027.4%9.56%
10011.7%17.4%
15017.4%22.9%
20025.6%26.9%
25023.3%32.5%
Download this series