I have some two-dimensional data (300 rows × 128k columns), and I need to sort each column. Is there an efficient method to do this?
You could use cub. cub::DeviceSegmentedRadixSort — CUB 104.0 documentation
An alternative to the segmented sort would be to call a separate sorting kernel per row. You would need to Benchmark both versions to see which is faster
1 Like
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.