d-v-b:feat/simd - Branch - zarr-developers/cast-value.rs

perf: add AVX2 SIMD via pulp, and benchmarks via criterion

#11Merged

Comparing

d-v-b:feat/simd

(

a9e4be6

) with

main

(

0c760a3

)

N/A

New: 45

Skipped: 21

Benchmarks

66 total

f32_to_u8/clamp[64000]

core/benches/conversions.rs::benches::bench_f32_to_u8::conversions

N/A

N/A190.2 µs

f64_to_u8/clamp[1024]

core/benches/conversions.rs::benches::bench_f64_to_u8::conversions

N/A

N/A6.3 µs

f32_to_u8/clamp[1000000]

core/benches/conversions.rs::benches::bench_f32_to_u8::conversions

N/A

N/A3 ms

i32_to_u8/clamp[1024]

core/benches/conversions.rs::benches::bench_i32_to_u8::conversions

N/A

N/A3.4 µs

f64_to_u8/clamp[1000000]

core/benches/conversions.rs::benches::bench_f64_to_u8::conversions

N/A

N/A5 ms

f64_to_u8/clamp[64000]

core/benches/conversions.rs::benches::bench_f64_to_u8::conversions

N/A

N/A322.5 µs

f64_to_u8/wrap[1024]

core/benches/conversions.rs::benches::bench_f64_to_u8::conversions

N/A

N/A22.6 µs

f32_to_u8/clamp[1024]

core/benches/conversions.rs::benches::bench_f32_to_u8::conversions

N/A

N/A4.1 µs

i32_to_u8/clamp+scalar_map[64000]

core/benches/conversions.rs::benches::bench_i32_to_u8::conversions

N/A

N/A493 µs

i32_to_u8/clamp+scalar_map[1000000]

core/benches/conversions.rs::benches::bench_i32_to_u8::conversions

N/A

N/A7.7 ms

f64_to_f32/nearest_even[64000]

core/benches/conversions.rs::benches::bench_f64_to_f32::conversions

N/A

N/A688.3 µs

f64_to_f32/clamp[1000000]

core/benches/conversions.rs::benches::bench_f64_to_f32::conversions

N/A

N/A6.3 ms

f64_to_f32/nearest_even[1000000]

core/benches/conversions.rs::benches::bench_f64_to_f32::conversions

N/A

N/A10.7 ms

f64_to_u8/wrap[1000000]

core/benches/conversions.rs::benches::bench_f64_to_u8::conversions

N/A

N/A21.3 ms

f64_to_f32/nearest_even[1024]

core/benches/conversions.rs::benches::bench_f64_to_f32::conversions

N/A

N/A11.5 µs

f64_to_i32/clamp[1000000]

core/benches/conversions.rs::benches::bench_f64_to_i32::conversions

N/A

N/A7.3 ms

i32_to_u8/clamp[1000000]

core/benches/conversions.rs::benches::bench_i32_to_u8::conversions

N/A

N/A2.8 ms

i64_to_f32/towards_zero[1024]

core/benches/conversions.rs::benches::bench_i64_to_f32::conversions

N/A

N/A11.5 µs

f64_to_f32/clamp[64000]

core/benches/conversions.rs::benches::bench_f64_to_f32::conversions

N/A

N/A403.8 µs

i64_to_f32/towards_zero[64000]

core/benches/conversions.rs::benches::bench_i64_to_f32::conversions

N/A

N/A688.3 µs

i32_to_u8/clamp+scalar_map[1024]

core/benches/conversions.rs::benches::bench_i32_to_u8::conversions

N/A

N/A8.4 µs

i64_to_f32/towards_zero[1000000]

core/benches/conversions.rs::benches::bench_i64_to_f32::conversions

N/A

N/A10.7 ms

f64_to_f32/clamp[1024]

core/benches/conversions.rs::benches::bench_f64_to_f32::conversions

N/A

N/A6.9 µs

i64_to_f32/nearest_even[1024]

core/benches/conversions.rs::benches::bench_i64_to_f32::conversions

N/A

N/A7.5 µs

i32_to_u8/clamp[64000]

core/benches/conversions.rs::benches::bench_i32_to_u8::conversions

N/A

N/A179.6 µs

Commits

Click on a commit to change the comparison range

Base

main

0c760a3

×6.1

Merge branch 'main' of https://github.com/zarr-developers/cast-value.rs into feat/simd

9d9f4a1

1 day ago

by d-v-b

Home Terms Privacy Docs