achalmers/try-further - Branch - KittyCAD/ezpz

Precompute more linear algebra outside the Newton Gauss loop

#160

Comparing

achalmers/try-further

(

7efc444

) with

main

(

0a477c1

)

+3%

Untouched: 8

Skipped: 5

Benchmarks

Skipped (5)

Passed

massively_parallel[600]

kcl-ezpz/benches/solver_bench.rs::benches::solve_massive

+7%

18.7 ms17.4 ms

massively_parallel_overconstrained[600]

kcl-ezpz/benches/solver_bench.rs::benches::solve_massive_overconstrained

+7%

18.7 ms17.4 ms

massively_parallel_overconstrained[200]

kcl-ezpz/benches/solver_bench.rs::benches::solve_massive_overconstrained

+7%

18.6 ms17.4 ms

massively_parallel[200]

kcl-ezpz/benches/solver_bench.rs::benches::solve_massive

+7%

18.6 ms17.4 ms

solve two rectangles dependent

kcl-ezpz/benches/solver_bench.rs::benches::solve_two_rectangles_dependent

+4%

174.1 µs166.7 µs

solve_inconsistent

kcl-ezpz/benches/solver_bench.rs::benches::solve_inconsistent

-1%

77.8 µs78.3 µs

solve_two_rectangles

kcl-ezpz/benches/solver_bench.rs::benches::solve_two_rectangles

-2%

146.4 µs149.8 µs

solve_nonsquare

kcl-ezpz/benches/solver_bench.rs::benches::solve_nonsquare

-5%

64.3 µs67.3 µs

Commits

Click on a commit to change the comparison range

Base

main

0a477c1

-11.68%

Precompute more

182ffe1

1 month ago

by adamchalmers

+0.55%

Precompute more of the 'a' matrix.

f39de2a

1 month ago

by adamchalmers

+14.11%

Move more out

aa6b34a

1 month ago

by adamchalmers

+0.06%

Merge branch 'main' into achalmers/try-further

7efc444

1 month ago

by adamchalmers

Home Terms Privacy Docs