Add row-oriented byte encoder (vortex-row crate)
#7985Comparing
claude/row-pr3-kernels
(7021233
) withdevelop
(faf7e42
)+18%
- Improvement
- 4
- Untouched
- 1217
Unknown Walltime execution environment detected
Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data. For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.Benchmarks
1221 total
chunked_varbinview_canonical_into[(1000, 10)]
vortex-array/benches/chunk_array_builder.rs+22%
197.9 µs162 µs
chunked_varbinview_into_canonical[(1000, 10)]
vortex-array/benches/chunk_array_builder.rs+20%
211.2 µs175.8 µs
chunked_varbinview_opt_canonical_into[(1000, 10)]
vortex-array/benches/chunk_array_builder.rs+19%
224.8 µs188.6 µs
chunked_varbinview_into_canonical[(100, 100)]
vortex-array/benches/chunk_array_builder.rs+11%
358.4 µs323.5 µs
chunked_varbinview_opt_canonical_into[(100, 100)]
vortex-array/benches/chunk_array_builder.rs+9%
410.8 µs377.3 µs
null_count_run_end[(10000, 256, 0.01)]
encodings/runend/benches/run_end_null_count.rs+5%
3.5 µs3.3 µs
null_count_run_end[(10000, 1024, 0.01)]
encodings/runend/benches/run_end_null_count.rs+5%
3.5 µs3.3 µs
null_count_run_end[(100000, 1024, 0.01)]
encodings/runend/benches/run_end_null_count.rs+5%
3.5 µs3.3 µs
cuda/runend/i32_runlen_1000[100M]
vortex-cuda/benches/runend_cuda.rs::benches::benchmark_runend::cuda+4%
904.3 µs867.4 µs
true_count_vortex_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+4%
764.7 ns735.6 ns
decode_varbin[(1000, 32)]
vortex-array/benches/dict_compress.rs+3%
27.2 µs26.3 µs
iter_vortex_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
923.3 ns894.2 ns
true_count_vortex_buffer[1024]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
923.3 ns894.2 ns
value_vortex_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
952.2 ns923.1 ns
chunked_bool_canonical_into[(1000, 10)]
vortex-array/benches/chunk_array_builder.rs+3%
45 µs43.6 µs
true_count_vortex_buffer[2048]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
965 ns935.8 ns
cuda/runend/i32_runlen_10[100M]
vortex-cuda/benches/runend_cuda.rs::benches::benchmark_runend::cuda+3%
1.4 ms1.4 ms
chunked_dict_fsst_into_canonical[(1000, 100, 100)]
encodings/fsst/benches/chunked_dict_fsst_builder.rs+3%
14 ms13.6 ms
decode_varbinview[(1000, 8)]
vortex-array/benches/dict_compress.rs+3%
27.9 µs27.2 µs
chunked_dict_fsst_canonical_into[(1000, 100, 100)]
encodings/fsst/benches/chunked_dict_fsst_builder.rs+3%
14 ms13.6 ms
chunked_dict_fsst_canonical_into[(1000, 10, 100)]
encodings/fsst/benches/chunked_dict_fsst_builder.rs+3%
13.9 ms13.6 ms
chunked_dict_fsst_canonical_into[(1000, 1000, 100)]
encodings/fsst/benches/chunked_dict_fsst_builder.rs+3%
14.9 ms14.5 ms
append_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+2%
1.4 µs1.4 µs
encode_primitives[u8, (1000, 4)]
vortex-array/benches/dict_compress.rs+2%
65.5 µs64.3 µs
decode_primitives[f32, (1000, 2)]
vortex-array/benches/dict_compress.rs+2%
17.2 µs16.9 µs
Commits
Click on a commit to change the comparison rangeBase
develop
faf7e42
-16.58%
FoR and Delta row-encode kernels (vortex-fastlanes)
0430217
3 days ago
by claude+34.57%
Fix CI: regenerate public-api lock and break vortex-row publish cycle
7021233
3 days ago
by claude