Table 4 Comparison of running time (/ms) for point sets derived from 3D models on GTX 660M
From: CudaChain: an alternative algorithm for finding 2D convex hulls on the GPU
3D Model | Size | Qhull | CudaChain | Speedup | |||
|---|---|---|---|---|---|---|---|
Total | GPU | CPU | CPU(%) | ||||
Armadillo | 172k | 47 | 39.7 | 37.6 | 2.1 | 5.29 | 1.2 |
Angel | 237k | 51 | 41.6 | 38.7 | 2.9 | 6.97 | 1.2 |
Skeleton hand | 327k | 77 | 45.4 | 41.5 | 3.9 | 8.59 | 1.7 |
Dragon | 437k | 98 | 59.8 | 53.9 | 5.9 | 9.87 | 1.6 |
Happy Buddha | 543k | 123 | 68.7 | 59.6 | 9.1 | 13.25 | 1.8 |
Turbine blade | 882k | 202 | 73.9 | 67.5 | 6.4 | 8.66 | 2.7 |
Vellum manuscript | 2M | 392 | 90.6 | 86.9 | 3.7 | 4.08 | 4.3 |
Asian dragon | 3M | 492 | 101.7 | 97.9 | 3.8 | 3.74 | 4.8 |
Thai statue | 5M | 547 | 106.0 | 102.4 | 3.6 | 3.40 | 5.2 |
Lucy | 14M | 1481 | 245.2 | 240.9 | 4.3 | 1.75 | 6.0 |