Skip to content

Conversation

@stephenswat
Copy link
Member

This commit reworks the covariance estimation logic to match that used in ACTS. It greatly increases the covariances of tracks, hopefully increasing the tracking efficiency.

@stephenswat stephenswat added the physics Physics performance monitoring label Oct 16, 2025
@stephenswat stephenswat force-pushed the phys/param_estimator_covariance branch 3 times, most recently from 562d024 to 9f6bb16 Compare October 17, 2025 09:16
This commit reworks the covariance estimation logic to match that used
in ACTS. It greatly increases the covariances of tracks, hopefully
increasing the tracking efficiency.
@stephenswat stephenswat force-pushed the phys/param_estimator_covariance branch from 9f6bb16 to 8552bc0 Compare October 28, 2025 14:44
@sonarqubecloud
Copy link

@stephenswat
Copy link
Member Author

Physics performance summary

Here is a summary of the physics performance effects of this PR. Command used:

traccc_seeding_example_cuda --input-directory=/data/Acts/odd-simulations-20240506/geant4_ttbar_mu200 --digitization-file=geometries/odd/odd-digi-geometric-config.json --detector-file=geometries/odd/odd-detray_geometry_detray.json --grid-file=geometries/odd/odd-detray_surface_grids_detray.json --material-file=geometries/odd/odd-detray_material_detray.json --input-events=10 --use-acts-geom-source=on --check-performance --truth-finding-min-track-candidates=5 --truth-finding-min-pt=1.0 --truth-finding-min-z=-150 --truth-finding-max-z=150 --truth-finding-max-r=10 --seed-matching-ratio=0.99 --track-matching-ratio=0.5 --track-candidates-range=5:100 --seedfinder-vertex-range=-150:150

Seeding performance

Total number of seeds went from 298345 to 298342 (-0.0%)

Seeding plots



Track finding performance

Total number of found tracks went from 63306 to 63628 (+0.5%)

Finding plots









Track fitting performance

Total number of fitted tracks went from 63306 to 63628 (+0.5%)

Fitting plots












Note

This is an automated message produced on the explicit request of a human being.

@stephenswat stephenswat marked this pull request as draft October 28, 2025 14:51
@stephenswat
Copy link
Member Author

Performance summary

Here is a summary of the performance effects of this PR:

Graphical

Tabular

KernelReciprocal ThroughputParallelism
7a195798552bc0Delta7a195798552bc0
propagate_to_next_surface9.83 ms9.42 ms-4.2%2.892.97
fit_forward3.05 ms3.70 ms21.1%5.214.85
fit_backward1.63 ms1.68 ms3.0%3.833.65
find_tracks1.25 ms1.30 ms4.2%1.771.75
ccl_kernel830.92 μs831.00 μs0.0%1.371.37
Thrust::sort728.84 μs728.62 μs-0.0%3.853.84
count_doublets629.78 μs629.95 μs0.0%1.611.61
count_triplets589.48 μs590.72 μs0.2%1.021.02
find_doublets453.65 μs442.08 μs-2.5%3.083.08
find_triplets172.18 μs172.94 μs0.4%1.321.31
select_seeds53.10 μs52.99 μs-0.2%1.341.34
remove_duplicates38.98 μs36.98 μs-5.1%16.6017.31
populate_grid23.33 μs23.40 μs0.3%1.221.22
count_grid_capacities22.09 μs22.13 μs0.2%1.221.22
fit_prelude16.55 μs17.26 μs4.3%15.2414.54
apply_interaction17.01 μs16.70 μs-1.8%5.805.86
update_triplet_weights15.03 μs15.10 μs0.4%1.271.27
estimate_track_params14.34 μs14.42 μs0.6%2.152.15
form_spacepoints13.14 μs13.17 μs0.2%1.481.48
fill_finding_propagation_sort_keys11.02 μs10.77 μs-2.3%6.326.42
build_tracks6.51 μs6.78 μs4.1%10.259.81
reduce_triplet_counts6.28 μs6.30 μs0.4%3.083.08
fill_finding_duplicate_removal_sort_keys3.33 μs3.19 μs-4.3%22.3923.18
unknown2.06 μs2.07 μs0.3%9.859.85
fill_fitting_sort_keys225.64 ns238.66 ns5.8%15.3714.68
fill_prefix_sum171.94 ns171.89 ns-0.0%341.30341.30
Total19.41 ms19.74 ms1.7%3.173.19

Important

All metrics in this report are given as reciprocal throughput, not as wallclock runtime.

Warning

At least one kernel incurred a significant performance regression.

Note

This is an automated message produced upon the explicit request of a human being.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

physics Physics performance monitoring

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants