Comparing
os/fbs-cuda
(2ea6a11
) withdevelop
(91ef114
)+15%
- Improvement
- 1
- Untouched
- 862
- Archived
- 1549
Benchmarks
2412 total
encode_varbin[(10000, 2)]
vortex-array/benches/dict_compress.rs+15%
1.6 ms1.4 ms
map_each[BufferMut<i32>, 128]
vortex-buffer/benches/vortex_buffer.rs+7%
887.2 ns828.9 ns
bench_dict_mask[(0.5, 0.9)]
vortex-array/benches/dict_mask.rs+7%
1.9 ms1.7 ms
binary_search_vortex
vortex-array/benches/search_sorted.rs-6%
880.6 ns938.9 ns
bench_dict_mask[(0.01, 0.9)]
vortex-array/benches/dict_mask.rs-6%
1.7 ms1.9 ms
bench_dict_mask[(0.1, 0.5)]
vortex-array/benches/dict_mask.rs-6%
1.9 ms2 ms
compress[(1000, 16)]
encodings/runend/benches/run_end_compress.rs-5%
62.5 µs66.1 µs
bench_dict_mask[(0.01, 0.5)]
vortex-array/benches/dict_mask.rs-5%
1.9 ms2 ms
compress[(1000, 256)]
encodings/runend/benches/run_end_compress.rs-5%
58.5 µs61.8 µs
bench_dict_mask[(0.5, 0.01)]
vortex-array/benches/dict_mask.rs-5%
2.1 ms2.3 ms
bench_dict_mask[(0.1, 0.1)]
vortex-array/benches/dict_mask.rs-5%
2.1 ms2.2 ms
bench_dict_mask[(0.5, 0.1)]
vortex-array/benches/dict_mask.rs-5%
2.1 ms2.2 ms
bench_dict_mask[(0.01, 0.1)]
vortex-array/benches/dict_mask.rs-5%
2.1 ms2.2 ms
bench_dict_mask[(0.1, 0.01)]
vortex-array/benches/dict_mask.rs-5%
2.2 ms2.3 ms
bench_dict_mask[(0.01, 0.01)]
vortex-array/benches/dict_mask.rs-5%
2.2 ms2.3 ms
take_map[(0.005, 1.0)]
vortex-array/benches/take_patches.rs+5%
2.5 ms2.4 ms
set_indices_vortex_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+5%
1.3 µs1.3 µs
binary_search_std
vortex-array/benches/search_sorted.rs-4%
653.3 ns682.5 ns
compress[(1000, 4)]
encodings/runend/benches/run_end_compress.rs-4%
77 µs80.1 µs
iter_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+4%
1.6 µs1.5 µs
decompress[u8, (1000, 4)]
encodings/runend/benches/run_end_compress.rs+4%
26 µs25.1 µs
true_count_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
917.8 ns888.6 ns
decode_primitives[f32, (1000, 2)]
vortex-array/benches/dict_compress.rs+3%
24.5 µs23.7 µs
null_count_run_end[(10000, 256, 0.01)]
encodings/runend/benches/run_end_null_count.rs-3%
3.7 µs3.8 µs
null_count_run_end[(100000, 1024, 0.01)]
encodings/runend/benches/run_end_null_count.rs-3%
3.7 µs3.8 µs
Commits
Click on a commit to change the comparison rangeBase
develop
91ef114
+0.01%
do not double validate the flatbuffer on the cuda layout
f40a7df
4 months ago
by onursatici+14.87%
lock
2ea6a11
4 months ago
by onursatici