Comparing
development
(28ee696
) withmain
(b7d3466
)-86%
- Improvement
- 30
- Regression
- 40
- Untouched
- 69
Different runtime environments detected
Some benchmarks with significant performance changes were compared across different runtime environments, which may affect the accuracy of the results. Learn more about regression causes.Benchmarks
139 total
precompute_only
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::cached_evaluator_and_precompute_runtime_benchmarks::precompute_stage_only-86%
1.5 ms11 ms
expression_value_eval_only
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::cached_evaluator_and_precompute_runtime_benchmarks::stage_isolated_cached_value_and_expression-33%
6.1 µs9.2 µs
projection_total_generated_mc_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-26%
528.1 µs717.6 µs
expression_gradient_eval_only
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::cached_evaluator_and_precompute_runtime_benchmarks::stage_isolated_cached_gradient_and_expression-22%
12 µs15.4 µs
projection_total_generated_mc_with_gradient_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-22%
2.7 ms3.5 ms
nll_value_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-21%
889.3 µs1,131.1 µs
projection_total_with_gradient_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-20%
2.7 ms3.3 ms
evaluate_value_small_sample
crates/laddu-core/benches/expression_ir_benchmarks.rs::benches::expression_runtime_benchmarks::expression_runtime_paths-19%
89.4 µs110.8 µs
projection_total_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-19%
460.8 µs570.9 µs
evaluate_local_real_unary[real_unary_heavy@4096]
crates/laddu-core/benches/expression_ir_benchmarks.rs::benches::expression_runtime_benchmarks::expression_ir_instruction_mix_generic-18%
1.2 ms1.4 ms
weighted_value_sum/n4k[real_unary_heavy@4096]
crates/laddu-core/benches/expression_ir_benchmarks.rs::benches::expression_ir_normalization_factorization_benchmarks::expression_ir_instruction_mix_real_unary_normalization-18%
1.3 ms1.6 ms
weighted_value_sum/n32k[real_unary_heavy@32768]
crates/laddu-core/benches/expression_ir_benchmarks.rs::benches::expression_ir_normalization_factorization_benchmarks::expression_ir_instruction_mix_real_unary_normalization-18%
10.1 ms12.3 ms
single_bin_nll_value_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_single_bin-16%
171.7 µs205.1 µs
nll_value_and_gradient_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-15%
5.6 ms6.6 ms
repeated_single_subset_calls[1]
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::projection_subset_cardinality_small_sample-14%
464.5 µs542.7 µs
repeated_single_subset_calls[2]
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::projection_subset_cardinality_small_sample-14%
854.2 µs996.5 µs
batched_subsets_single_call[2]
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::projection_subset_cardinality_small_sample-14%
854.1 µs995.6 µs
batched_subsets_single_call[4]
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::projection_subset_cardinality_small_sample-14%
1.4 ms1.6 ms
evaluate_local/n4k[separable@4096]
crates/laddu-core/benches/expression_ir_benchmarks.rs::benches::expression_ir_normalization_factorization_benchmarks::expression_ir_factorization_controls-14%
804 µs936.2 µs
repeated_single_subset_calls[4]
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::projection_subset_cardinality_small_sample-14%
1.4 ms1.6 ms
evaluate_local/n32k[separable@32768]
crates/laddu-core/benches/expression_ir_benchmarks.rs::benches::expression_ir_normalization_factorization_benchmarks::expression_ir_factorization_controls-14%
6.2 ms7.1 ms
projection_component_s_wave_with_gradient_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-14%
1.6 ms1.9 ms
batched_subsets_single_call[1]
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::projection_subset_cardinality_small_sample-13%
460.3 µs531.6 µs
nll_gradient_small_sample
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::breit_wigner_partial_wave_unbinned-13%
4.8 ms5.5 ms
batched_subsets_single_call[8]
crates/laddu/benches/workflow_behavior_cpu_benchmarks.rs::benches::breit_wigner_partial_wave_benchmarks::projection_subset_cardinality_small_sample-12%
2.7 ms3.1 ms
Commits
Click on a commit to change the comparison rangeBase
main
b7d3466
-86.03%
fix: some clippy lints and a better orthogonality check
6a23fb3
10 days ago
by denehoffman0%
fix: correct floating-point error in resampling test
18008cc
10 days ago
by denehoffman-0.07%
perf: read lowered program only once per call rather than once per event
aa17280
10 days ago
by denehoffman0%
perf: remove a few more RwLock reads inside loops
28ee696
10 days ago
by denehoffman