Implementation of parallel stream compaction using Numba CUDA.
| N | gpu time (ms) | cpu time (ms) |
|---|---|---|
| 10^4 | 1.20 | 0.362 |
| 10^5 | 1.99 | 3.40 |
| 10^6 | 3.49 | 37.0 |
| 10^7 | 13.3 | 356.0 |
| 10^8 | 70.8 | 3533 |
| 10^9 | 673 | 35934 |
^ tested on machine with Ryzen 5 3600 and RTX 3090.
Based on the following paper:
Hughes, D. et al. (2013). InK‐Compact: In‐Kernel Stream Compaction and Its Application to Multi‐Kernel Data Visualization on General‐Purpose GPUs.