Benchmarks of ordito.holes against every library that computes the same thing. Each chart shows the time of every library relative to ordito on CUDA (1x, in green): a longer bar took longer. The table under each chart lists every input the function was measured on. See Benchmarks for how to read and reproduce these numbers.
Time relative to ordito (CUDA), geometric mean over the holes functions both implement; in parentheses, the number of functions. Log scale. Runs stopped at the time limit are left out.Values
Library
Time relative to ordito
Functions compared
ordito faster in
Open3D
3.9x
1
1
trimesh
4.8x
1
1
MeshLib
18x
11
11
MeshLab
25x
1
1
PyMeshFix
1226x
1
1
fill_cone
rim_long (131 072 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
11.85 ms
1x
MeshLib
81.37 ms
6.9x
All inputs
Input
ordito (CUDA)
MeshLib
holes_many
390.7 µs
12.55 ms
sphere_med
194.9 µs
10.74 ms
rim_long
11.85 ms
81.37 ms
fill_min_weight
holes_many (81 408 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
738.8 µs
1x
MeshLib
11.00 ms
15x
MeshLab
38.10 ms
52x
trimesh
42.82 ms
58x
Open3D
46.91 ms
63x
All inputs
Input
ordito (CUDA)
MeshLab
MeshLib
Open3D
trimesh
rim_short
4.88 ms
60.17 ms
74.10 ms
1.16 ms
1.94 ms
holes_many
738.8 µs
38.10 ms
11.00 ms
46.91 ms
42.82 ms
fill_smooth
holes_many (81 408 faces) · triangulate_only=False · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
3.18 ms
1x
MeshLib
4.643 s
1459x
All inputs
Input
ordito (CUDA)
MeshLib
rim_short triangulate_only=False
16.85 ms
36.94 ms
rim_short triangulate_only=True
4.90 ms
holes_many triangulate_only=False
3.18 ms
4.643 s
holes_many triangulate_only=True
845.2 µs
refill_region
dragon (871 414 faces) · triangulate_only=True · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
20.94 ms
1x
MeshLib
97.83 ms
4.7x
All inputs
Input
ordito (CUDA)
MeshLib
sphere_small triangulate_only=True
1.46 ms
sphere_small triangulate_only=False
13.02 ms
bunny_decimated triangulate_only=True
1.07 ms
11.77 ms
bunny_decimated triangulate_only=False
4.61 ms
18.83 ms
bunny triangulate_only=True
6.61 ms
17.69 ms
bunny triangulate_only=False
14.54 ms
30.39 ms
dragon triangulate_only=True
20.94 ms
97.83 ms
dragon triangulate_only=False
32.47 ms
161.98 ms
happy_buddha triangulate_only=True
47.21 ms
140.16 ms
happy_buddha triangulate_only=False
59.36 ms
186.66 ms
lucy triangulate_only=True
173.59 ms
3.755 s
lucy triangulate_only=False
435.52 ms
4.446 s
stitch_min_weight
rim_short (1 024 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
8.15 ms
1x
MeshLib
35.44 ms
4.3x
All inputs
Input
ordito (CUDA)
MeshLib
rim_short
8.15 ms
35.44 ms
fillable_loop_mask
dragon (871 414 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
335.2 µs
1x
MeshLib
857.5 µs
2.6x
All inputs
Input
ordito (CUDA)
MeshLib
saddle_small
124.2 µs
bunny_decimated
125.5 µs
100.2 µs
bunny
133.3 µs
183.6 µs
dragon
335.2 µs
857.5 µs
happy_buddha
1.70 ms
lucy
43.66 ms
extend_hole
dragon (871 414 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
339.2 µs
1x
MeshLib
55.93 ms
165x
All inputs
Input
ordito (CUDA)
MeshLib
saddle_small
133.5 µs
bunny_decimated
146.4 µs
8.00 ms
bunny
125.6 µs
9.08 ms
dragon
339.2 µs
55.93 ms
happy_buddha
68.75 ms
lucy
3.664 s
build_bottom
dragon (871 414 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
375.4 µs
1x
MeshLib
58.14 ms
155x
All inputs
Input
ordito (CUDA)
MeshLib
saddle_small
146.6 µs
bunny_decimated
154.1 µs
8.45 ms
bunny
145.7 µs
9.09 ms
dragon
375.4 µs
58.14 ms
happy_buddha
66.24 ms
lucy
3.432 s
bridge_edges
dragon (871 414 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
150.3 µs
1x
MeshLib
52.65 ms
350x
All inputs
Input
ordito (CUDA)
MeshLib
saddle_small
132.4 µs
bunny_decimated
158.9 µs
7.81 ms
bunny
134.0 µs
8.35 ms
dragon
150.3 µs
52.65 ms
lucy
1.02 ms
crashed (aborted)
bridge_edges_smooth
dragon (871 414 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
468.6 µs
1x
MeshLib
48.72 ms
104x
All inputs
Input
ordito (CUDA)
MeshLib
saddle_small
378.3 µs
bunny_decimated
360.3 µs
7.89 ms
bunny
340.8 µs
8.33 ms
dragon
468.6 µs
48.72 ms
lucy
1.52 ms
crashed (segfault)
join_closest_components
open_parts_4 (81 916 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
482.1 µs
1x
PyMeshFix
317.69 ms
659x
All inputs
Input
ordito (CUDA)
PyMeshFix
open_parts_64
1.31 ms
2.592 s
open_parts_16
569.5 µs
802.76 ms
open_parts_4
482.1 µs
317.69 ms
stitch_smooth
rim_short (1 024 faces) · time relative to ordito, log scaleValues
Library
Median
Relative to ordito
ordito (CUDA)
14.77 ms
1x
MeshLib
50.03 ms
3.4x
All inputs
Input
ordito (CUDA)
MeshLib
rim_short
14.77 ms
50.03 ms
Without a reference
No other library computes these, so they have no comparison. ordito (CUDA)'s median time on the largest input it was measured on: