Skip to content

perf: try out new cliping numba kernel #3952

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
ilan-gold wants to merge 9 commits into
base: main
Choose a base branch
from
+77 −35

fix: maybe?

8b3f328
Select commit
Loading
Failed to load commit list.
Draft

perf: try out new cliping numba kernel #3952

fix: maybe?
8b3f328
Select commit
Loading
Failed to load commit list.
scverse-benchmark / benchmark succeeded Jan 30, 2026 in 1h 1m 31s

Benchmark

Benchmark run successful

Details

All benchmarks:

Change Before [4c7c5c0] After [8b3f328] Ratio Benchmark (Parameter)
308M 308M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('bmmc', 'counts')
308M 307M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('bmmc', 'counts-off-axis')
4.1G 4.1G 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('lung93k', 'counts')
4.1G 4.1G 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('lung93k', 'counts-off-axis')
368M 367M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc3k', 'counts')
367M 367M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc3k', 'counts-off-axis')
286M 286M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc68k_reduced', 'counts')
285M 285M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc68k_reduced', 'counts-off-axis')
310M 310M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('bmmc', 'counts')
310M 310M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('bmmc', 'counts-off-axis')
4.44G 4.44G 1.00 preprocessing_counts.FastSuite.peakmem_log1p('lung93k', 'counts')
4.44G 4.44G 1.00 preprocessing_counts.FastSuite.peakmem_log1p('lung93k', 'counts-off-axis')
376M 376M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc3k', 'counts')
376M 376M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc3k', 'counts-off-axis')
285M 285M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc68k_reduced', 'counts')
285M 284M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc68k_reduced', 'counts-off-axis')
401M 396M 0.99 preprocessing_counts.FastSuite.peakmem_normalize_total('bmmc', 'counts')
396M 396M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('bmmc', 'counts-off-axis')
4.96G 4.96G 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('lung93k', 'counts')
4.96G 4.96G 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('lung93k', 'counts-off-axis')
471M 466M 0.99 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc3k', 'counts')
467M 466M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc3k', 'counts-off-axis')
285M 285M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc68k_reduced', 'counts')
285M 285M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc68k_reduced', 'counts-off-axis')
13.0±0.4ms 13.5±0.3ms 1.04 preprocessing_counts.FastSuite.time_calculate_qc_metrics('bmmc', 'counts')
12.6±0.4ms 12.6±0.3ms 1.00 preprocessing_counts.FastSuite.time_calculate_qc_metrics('bmmc', 'counts-off-axis')
2.15±0.01s 2.14±0.01s 1.00 preprocessing_counts.FastSuite.time_calculate_qc_metrics('lung93k', 'counts')
1.66±0s 1.67±0s 1.00 preprocessing_counts.FastSuite.time_calculate_qc_metrics('lung93k', 'counts-off-axis')
38.4±0.7ms 38.4±0.5ms 1.00 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc3k', 'counts')
29.2±0.7ms 28.3±0.5ms 0.97 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc3k', 'counts-off-axis')
5.00±0.02ms 4.95±0.04ms 0.99 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc68k_reduced', 'counts')
4.95±0.02ms 4.95±0.01ms 1.00 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc68k_reduced', 'counts-off-axis')
1.55±0.1ms 1.56±0.1ms 1.01 preprocessing_counts.FastSuite.time_log1p('bmmc', 'counts')
1.61±0.08ms 1.61±0.05ms 1.00 preprocessing_counts.FastSuite.time_log1p('bmmc', 'counts-off-axis')
692±7ms 692±6ms 1.00 preprocessing_counts.FastSuite.time_log1p('lung93k', 'counts')
690±0.7ms 690±4ms 1.00 preprocessing_counts.FastSuite.time_log1p('lung93k', 'counts-off-axis')
7.40±0.04ms 7.44±0.06ms 1.01 preprocessing_counts.FastSuite.time_log1p('pbmc3k', 'counts')
7.49±0.4ms 7.51±0.5ms 1.00 preprocessing_counts.FastSuite.time_log1p('pbmc3k', 'counts-off-axis')
393±4μs 390±1μs 0.99 preprocessing_counts.FastSuite.time_log1p('pbmc68k_reduced', 'counts')
392±4μs 391±5μs 1.00 preprocessing_counts.FastSuite.time_log1p('pbmc68k_reduced', 'counts-off-axis')
2.84±0.05ms 2.88±0.07ms 1.02 preprocessing_counts.FastSuite.time_normalize_total('bmmc', 'counts')
7.21±0.2ms 7.36±0.5ms 1.02 preprocessing_counts.FastSuite.time_normalize_total('bmmc', 'counts-off-axis')
588±0.8ms 600±4ms 1.02 preprocessing_counts.FastSuite.time_normalize_total('lung93k', 'counts')
2.79±0s 2.94±0.02s 1.05 preprocessing_counts.FastSuite.time_normalize_total('lung93k', 'counts-off-axis')
9.71±0.5ms 9.48±0.1ms 0.98 preprocessing_counts.FastSuite.time_normalize_total('pbmc3k', 'counts')
35.7±1ms 36.5±0.7ms 1.02 preprocessing_counts.FastSuite.time_normalize_total('pbmc3k', 'counts-off-axis')
597±80μs 577±40μs 0.97 preprocessing_counts.FastSuite.time_normalize_total('pbmc68k_reduced', 'counts')
552±2μs 550±2μs 1.00 preprocessing_counts.FastSuite.time_normalize_total('pbmc68k_reduced', 'counts-off-axis')
422M 422M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc3k', 'counts')
422M 422M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc3k', 'counts-off-axis')
296M 296M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc68k_reduced', 'counts')
296M 297M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc68k_reduced', 'counts-off-axis')
422M 422M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc3k', 'counts')
422M 422M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc3k', 'counts-off-axis')
296M 295M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc68k_reduced', 'counts')
296M 296M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc68k_reduced', 'counts-off-axis')
1.1G 1.1G 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc3k', 'counts')
1.1G 1.1G 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc3k', 'counts-off-axis')
513M 525M 1.02 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc68k_reduced', 'counts')
523M 514M 0.98 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc68k_reduced', 'counts-off-axis')
65.0±4ms 65.8±4ms 1.01 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc3k', 'counts')
89.6±0.7ms 89.8±1ms 1.00 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc3k', 'counts-off-axis')
9.34±0.6ms 10.4±0.9ms ~1.11 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc68k_reduced', 'counts')
10.3±0.8ms 10.9±0.6ms 1.06 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc68k_reduced', 'counts-off-axis')
59.7±4ms 59.3±3ms 0.99 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc3k', 'counts')
83.9±3ms 85.2±3ms 1.02 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc3k', 'counts-off-axis')
11.1±0.5ms 13.2±2ms ~1.18 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc68k_reduced', 'counts')
9.53±0.6ms 11.1±1ms ~1.16 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc68k_reduced', 'counts-off-axis')
2.71±0.3s 2.56±0.08s 0.94 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc3k', 'counts')
5.58±4s 2.57±0.2s ~0.46 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc3k', 'counts-off-axis')
549±8ms 548±20ms 1.00 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc68k_reduced', 'counts')
547±6ms 548±10ms 1.00 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc68k_reduced', 'counts-off-axis')
459M 458M 1.00 preprocessing_log.HVGSuite.peakmem_highly_variable_genes('cell_ranger')
458M 458M 1.00 preprocessing_log.HVGSuite.peakmem_highly_variable_genes('seurat')
459M 468M 1.02 preprocessing_log.HVGSuite.peakmem_highly_variable_genes('seurat_v3')
31.5±0.4ms 31.4±0.4ms 1.00 preprocessing_log.HVGSuite.time_highly_variable_genes('cell_ranger')
35.1±0.2ms 35.2±0.2ms 1.00 preprocessing_log.HVGSuite.time_highly_variable_genes('seurat')
- 96.8±0.3ms 80.0±0.3ms 0.83 preprocessing_log.HVGSuite.time_highly_variable_genes('seurat_v3')
560M 561M 1.00 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc3k', 'off-axis')
583M 578M 0.99 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc3k', None)
488M 493M 1.01 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc68k_reduced', 'off-axis')
491M 495M 1.01 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc68k_reduced', None)
n/a n/a n/a preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc3k', 'off-axis')
n/a n/a n/a preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc3k', None)
347M 344M 0.99 preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc68k_reduced', 'off-axis')
348M 348M 1.00 preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc68k_reduced', None)
1.29G 1.29G 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc3k', 'off-axis')
1.48G 1.48G 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc3k', None)
337M 336M 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc68k_reduced', 'off-axis')
334M 334M 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc68k_reduced', None)
1.93±0.02s 1.95±0.01s 1.01 preprocessing_log.PreprocessingSuite.time_pca('pbmc3k', 'off-axis')
2.14±0.01s 2.15±0.01s 1.00 preprocessing_log.PreprocessingSuite.time_pca('pbmc3k', None)
180±10ms 107±60ms ~0.59 preprocessing_log.PreprocessingSuite.time_pca('pbmc68k_reduced', 'off-axis')
168±40ms 160±30ms 0.95 preprocessing_log.PreprocessingSuite.time_pca('pbmc68k_reduced', None)
n/a n/a n/a preprocessing_log.PreprocessingSuite.time_regress_out('pbmc3k', 'off-axis')
n/a n/a n/a preprocessing_log.PreprocessingSuite.time_regress_out('pbmc3k', None)
17.9±1ms 16.8±1ms 0.94 preprocessing_log.PreprocessingSuite.time_regress_out('pbmc68k_reduced', 'off-axis')
16.5±0.4ms 17.1±0.8ms 1.04 preprocessing_log.PreprocessingSuite.time_regress_out('pbmc68k_reduced', None)
504±3ms 496±1ms 0.98 preprocessing_log.PreprocessingSuite.time_scale('pbmc3k', 'off-axis')
560±2ms 552±3ms 0.99 preprocessing_log.PreprocessingSuite.time_scale('pbmc3k', None)
4.80±0.3ms 5.02±0.4ms 1.05 preprocessing_log.PreprocessingSuite.time_scale('pbmc68k_reduced', 'off-axis')
4.73±0.08ms 4.72±0.1ms 1.00 preprocessing_log.PreprocessingSuite.time_scale('pbmc68k_reduced', None)
285M 285M 1.00 tools.ToolsSuite.peakmem_diffmap
291M 291M 1.00 tools.ToolsSuite.peakmem_leiden
362M 362M 1.00 tools.ToolsSuite.peakmem_rank_genes_groups
444M 444M 1.00 tools.ToolsSuite.peakmem_umap
15.2±0.3ms 14.4±0.4ms 0.95 tools.ToolsSuite.time_diffmap
18.6±0.4ms 18.0±0.3ms 0.97 tools.ToolsSuite.time_leiden
62.2±0.5ms 62.8±0.5ms 1.01 tools.ToolsSuite.time_rank_genes_groups
1.02±0s 1.02±0s 1.00 tools.ToolsSuite.time_umap
View more details on scverse-benchmark