Skip to content

Sum Biot-Savart over coils with lax.scan in one shared kernel - #208

Open
rogeriojorge wants to merge 5 commits into
mainfrom
perf/biot-savart-scan
Open

rogeriojorge wants to merge 5 commits into
mainfrom
perf/biot-savart-scan

Conversation

@rogeriojorge

@rogeriojorge rogeriojorge commented Nov 10, 2026 •

Copy link
Copy Markdown
Member

Reduce peak memory in the Biot–Savart evaluation. The earlier GPU benchmark used 4–15 times less device memory. Field evaluation with many coils can be up to 25% slower.

For 16 coils on the laptop CPU, the warm field evaluation took 13.4 ms before and 14.0 ms after. The field derivative improved from 33.7 to 23.9 ms. Timing and memory by coil count

Measurements and test conditions

@rogeriojorge
rogeriojorge marked this pull request as ready for review November 11, 2026 03:47
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant