Comparing
ji/inline-kernel
(bb5f86c
) withdevelop
(8a25a39
)-90%
- Improvements
- 0
- Regressions
- 10
- Untouched
- 10
- New
- 1
- Dropped
- 0
- Ignored
- 0
Benchmarks
Failed
pack 16 -> 3 stackRegression
benches/bitpacking.rs::benches::pack::pack::pack 16 -> 3 stack-31%
1.2 µs
1.7 µs
unpack single 16 <- 3Regression
benches/bitpacking.rs::benches::pack::unpack-single::unpack single 16 <- 3-12%
8.4 µs
9.6 µs
unpack 16 <- 3 stackRegression
benches/bitpacking.rs::benches::pack::unpack::unpack 16 <- 3 stack-90%
1.1 µs
10.3 µs
bitpacking_cmp_fused[u16]Regression
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_fused[u16]-12%
1.3 µs
1.5 µs
bitpacking_cmp_seq[u32]Regression
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_seq[u32]-17%
4.8 µs
5.8 µs
bitpacking_cmp_seq[u64]Regression
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_seq[u64]-24%
6.3 µs
8.3 µs
bitpacking_cmp_unpack[u16]Regression
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_unpack[u16]-34%
1.1 µs
1.7 µs
bitpacking_cmp_unpack[u32]Regression
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_unpack[u32]-35%
1.9 µs
2.8 µs
bitpacking_cmp_unpack[u64]Regression
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_unpack[u64]-39%
3.2 µs
5.2 µs
delta u16 fusedRegression
benches/delta.rs::benches::delta::delta::delta u16 fused-11%
2.1 µs
2.3 µs
New
unchecked unpack 16 <- 3 stack
benches/bitpacking.rs::benches::pack::unpack::unchecked unpack 16 <- 3 stackN/A
N/A
10.5 µs
Passed
transpose u16
benches/transpose.rs::benches::transpose::transpose::transpose u160%
9.5 µs
9.5 µs
decompress
benches/delta.rs::benches::throughput::throughput::decompress0%
1.2 ms
1.2 ms
compress
benches/delta.rs::benches::throughput::throughput::compress0%
1.2 ms
1.3 ms
decompress
benches/bitpacking.rs::benches::throughput::throughput::decompress0%
1.2 ms
1.2 ms
compress
benches/bitpacking.rs::benches::throughput::throughput::compress0%
1.2 ms
1.3 ms
bitpacking_cmp_fused[u64]
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_fused[u64]-3%
3.1 µs
3.2 µs
pack 16 -> 3 heap
benches/bitpacking.rs::benches::pack::pack::pack 16 -> 3 heap-3%
1.7 µs
1.7 µs
bitpacking_cmp_fused[u32]
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_fused[u32]-4%
1.9 µs
2 µs
delta u16 unfused
benches/delta.rs::benches::delta::delta::delta u16 unfused-6%
3.6 µs
3.8 µs
bitpacking_cmp_seq[u16]
benches/bitpacking_cmp_cod.rs::bitpacking_cmp_seq[u16]-9%
5.1 µs
5.6 µs
Commits
Click on a commit to change the comparison rangeBase
develop
8a25a39
-39%
never inline kernel impls to improve compile times
b44f302
2 months ago
by joseph-isaacs-51%
update the bench
bb5f86c
2 months ago
by joseph-isaacs