[experimental] metal execution
#6868Comparing
metal
(cf80187
) withdevelop
(1bb7e68
)0%
- Untouched
- 1000
- New
- 12
- Skipped
- 1466
Benchmarks
2478 total
for_decompress_i32[(1000000, "1M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A4.1 ms
for_decompress_i32[(10000000, "10M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A58 ms
for_decompress_i64[(1000000, "1M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A11.4 ms
for_decompress_u32[(1000000, "1M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A4.1 ms
for_decompress_i32[(100000, "100K")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A742.1 µs
for_decompress_u64[(10000000, "10M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A116 ms
for_decompress_u64[(100000, "100K")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A1.5 ms
for_decompress_i64[(10000000, "10M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A116 ms
for_decompress_u64[(1000000, "1M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A11.4 ms
for_decompress_u32[(100000, "100K")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A742.1 µs
for_decompress_u32[(10000000, "10M")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A58 ms
for_decompress_i64[(100000, "100K")]
encodings/fastlanes/benches/for_decompression.rsN/A
N/A1.5 ms
encode_varbin[(10000, 8)]
vortex-array/benches/dict_compress.rs+4%
1.5 ms*1.4 ms
set_indices_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
1.2 µs1.1 µs
take_10_contiguous
encodings/fastlanes/benches/bitpacking_take.rs+2%
30.3 µs*29.7 µs
chunked_dict_primitive_into_canonical[f32, (1000, 10, 100)]
vortex-array/benches/chunked_dict_builder.rs+2%
976.6 µs*958.4 µs
chunked_dict_primitive_canonical_into[u32, (1000, 1000, 10)]
vortex-array/benches/chunked_dict_builder.rs+2%
137.7 µs*135.4 µs
chunked_dict_primitive_into_canonical[f64, (1000, 100, 10)]
vortex-array/benches/chunked_dict_builder.rs+1%
171.8 µs*169.5 µs
chunked_dict_primitive_into_canonical[f32, (1000, 10, 10)]
vortex-array/benches/chunked_dict_builder.rs+1%
127.1 µs*125.4 µs
chunked_dict_primitive_into_canonical[u64, (1000, 10, 10)]
vortex-array/benches/chunked_dict_builder.rs+1%
168.3 µs*166.1 µs
chunked_dict_primitive_into_canonical[f32, (1000, 100, 10)]
vortex-array/benches/chunked_dict_builder.rs+1%
128.6 µs*127.1 µs
chunked_dict_primitive_canonical_into[u64, (1000, 100, 10)]
vortex-array/benches/chunked_dict_builder.rs+1%
163.6 µs*161.6 µs
chunked_dict_primitive_canonical_into[f64, (1000, 10, 100)]
vortex-array/benches/chunked_dict_builder.rs+1%
1.3 ms*1.3 ms
push_n_vortex_buffer[u32, 128]
vortex-buffer/benches/vortex_buffer.rs+1%
2.9 µs2.9 µs
i32_small_overlapping
vortex-array/benches/listview_rebuild.rs+1%
70.4 µs*69.7 µs
Commits
Click on a commit to change the comparison rangeBase
develop
1bb7e68
-0.07%
feat(vortex-metal): improve benchmark with preloaded device buffers
cf80187
14 days ago
by a10y