Optimize CUDA cluster finder transfers and kernel hot path

Browse Source

- Use per-stream pinned host staging buffers for truly async CUDA transfers.
- Avoid reserving full device capacity per result frame.
- Reduce kernel work by delaying cluster payload construction.
- Use squared comparisons and removing per-pixel sqrtf() ops.

...

This commit is contained in:

kferjaoui

2026-04-30 18:23:31 +02:00

parent 34e69a8065

commit 88e0e8d678

3 changed files with 241 additions and 91 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

python/tests/ClusterFinderCUDA.ipynb

+105 -36

File diff suppressed because one or more lines are too long