Internal:Metric Adaptive CDT3D: Difference between revisions
Jump to navigation
Jump to search
add stability data |
|||
| Line 278: | Line 278: | ||
[[File:Histogram_ratio_reconstruct.png|800px]] | [[File:Histogram_ratio_reconstruct.png|800px]] | ||
==Comparison Surface Adaptation vs Mesh Reconstruction == | ==Comparison Surface Adaptation vs Mesh Reconstruction (reproducibility) == | ||
[[File:Histogram_mean_compare.png|800px]] | [[File:Histogram_mean_compare.png|800px]] | ||
[[File:Histogram_ratio_compare.png|800px]] | [[File:Histogram_ratio_compare.png|800px]] | ||
Surface | Surface adaptation offer better edge lengths but mesh reconstruction slightly better mean ratio quality | ||
Revision as of 10:59, 16 May 2018
Delta Wing (Scitech 2018)
50k Complexity
Starting from a coarse surface mesh
| # Threads | # Tetrahedra | # Vertices | Intitialize Bg Mesh | Boundary Adaptation (MMGS) | Boundary Recovery | Mesh Refinement | Total Time
(without MMGS) |
Memory (GB) | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| # Iterations | Par. Deactivation | Par. Vertex Creation | Vertex Insertion | Partition | Par. Reconnection | ||||||||
| 1 | 357,165 | 62,177 | 1.44 | 6.32 | 0.36 | 20 | 2.09 | 11.92 | 0.09 | 1.26 | 57.35 | 80.97 | 0.44 |
| 6 | 361,037 | 62,795 | 1.49 | 6.23 | 0.35 | 21 | 0.51 | 2.78 | 0.11 | 1.45 | 14.06 | 27.12 | 0.63 |
| 12 | 363,278 | 63,198 | 1.41 | 6.27 | 0.33 | 22 | 0.32 | 1.55 | 0.11 | 1.62 | 8.24 | 19.99 | 0.86 |
| 24 | 363,756 | 63,270 | 1.49 | 6.26 | 0.33 | 22 | 0.2 | 0.88 | 0.11 | 1.6 | 5.13 | 16.15 | 1.31 |
Notes 1
- according to Mike's email mesh should be around 100k vertices
- Metric Input Data available at https://github.com/UGAWG/parallel-adapt-results
- Input surface is the coase mesh available at https://github.com/UGAWG/solution-adapt-cases
- The experiments performed on a DELL workstation with Linux Red Hat Enterprise, 24 hardware cores Intel® Xeon® CPU E5-2697v2@2.70 GHz, and 757 GB RAM
- Memory is measured at the end of Mesh refinement and thus it may reach a slightly higher value before the point of measurement
- parameters
--brec=2 --met=1 --meval=3 --mrecm2=1 --mrec4=1 --nadapt=2 --malign=0 --mpp=2 --ndfs=0 --nthreads=$THREADS --nbuckets=20*$THREADS --nthreadsaf=20*$THREADS --nbucketsaf=20*${THREADS} --blockmul=500000 --nqual=0 --nminpts=30 --ordt=0 --pardeact=1 --mdse=0 - using commit
bb7819619eb9b8831ac1d4515b9605869df72b47or tagDelta-wiki
500k Complexity
Starting from a 50K Complexity volume mesh
Size of intial mesh Vertices : 104,305 Tetrahedra: 581,196
| # Threads | # Tetrahedra | # Vertices | Intitialize Bg Mesh | Reconstruct Mesh | Mesh Refinement | Total Time | Memory (GB) | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| # Iterations | Par. Deactivation | Par. Vertex Creation | Vertex Insertion | Partition | Par. Reconnection | |||||||
| 1 | 3,079,286 | 521,152 | 1.18 | 2.02 | 31 | 34.66 | 138.72 | 0.82 | 26.6 | 1115.09 | 1320.18 | 1.34 |
| 6 | 3,124,862 | 528,775 | 1.48 | 2.03 | 34 | 9.44 | 36.18 | 0.82 | 33.69 | 184.91 | 269.65 | 1.35 |
| 12 | 3,136,586 | 530,696 | 1.15 | 2.03 | 32 | 4.92 | 19.09 | 0.82 | 33.1 | 101.82 | 164.04 | 1.42 |
| 24 | 3,147,672 | 532,467 | 1.17 | 2.03 | 30 | 2.63 | 10.34 | 1.03 | 37.51 | 56.45 | 112.26 | 1.76 |
This approach offers no boundary adaptation.
Starting for a coarse surface mesh
| # Threads | # Tetrahedra | # Vertices | Intitialize Bg Mesh | Boundary Adaptation (MMGS) | Boundary Recovery | Mesh Refinement | Total Time
(without MMGS) |
Memory (GB) | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| # Iterations | Par. Deactivation | Par. Vertex Creation | Vertex Insertion | Partition | Par. Reconnection | ||||||||
| 1 | 3,301,987 | 564,041 | 1.3 | 29.83 | 6.39 | 26 | 26.59 | 112.14 | 1.0 | 19.41 | 858.58 | 1,055.77 | 1.34 |
| 6 | 3,327,519 | 568,583 | 1.48 | 29.64 | 5.3 | 28 | 7.37 | 28.32 | 1.08 | 26.08 | 165.42 | 265.23 | 1.25 |
| 12 | 3,333,780 | 569,607 | 1.45 | 30.8 | 6.87 | 29 | 4.18 | 16.91 | 1.23 | 31.54 | 95.14 | 188.78 | 1.41 |
| 24 | 3,333,430 | 569,592 | 1.33 | 29.93 | 5.96 | 30 | 2.47 | 9.33 | 0.99 | 28.92 | 53.59 | 133.08 | 1.75 |
Notes 2
- TODO compare quality of two methods
- according to Mike's email mesh should be around 1M vertices
- in some cases boundary recovery reached a point of non-developed code. Rerunning created a different surface mesh and CDT3D was able to continue.
- commit
912b8d5ec5b171ebf20bc6a61b9f7fe572f3d3d5 - parameters
--met=1 --meval=3 --mrecm2=1 --mrec4=1 --nadapt=1 --reconstruct=1 --malign=0 --mpp=2 --qstatr=1 --outmesh=${NAME}.meshb --ndfs=0 --nthreads=$THREADS --nbuckets=20*NTHREADS --nthreadsaf=20*NTHREADS --nbucketsaf=20*NTHREADS --blockmul=500000 --nqual=0 --nminpts=30 --ordt=0 --pardeact=1 --mdse=0
General Notes
- smoothing : Tetrahedral Mesh Improvement Using Swapping and Smoothing, Lori A. Freitag and Carl Ollivier-Gooch PDF
- An Efficient Method to Solve the Minmax Problem Directly PDF
Stability for 500k case
Comparison of Quality Metric using different amount of cores and surface adaptation
Comparison of Quality Metric using different amount of cores and mesh reconstruction
Comparison Surface Adaptation vs Mesh Reconstruction (reproducibility)
Surface adaptation offer better edge lengths but mesh reconstruction slightly better mean ratio quality