TurboQuant: Block Decomposition
#8139Comparing
ct/tq-block
(962c7fc
) withdevelop
(73454db
)-15%
- Regression
- 1
- Untouched
- 1274
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
1275 total
cuda/bitpacked_u8/unpack/3bw[100M]
vortex-cuda/benches/bitpacked_cuda.rs::benches::benchmark_bitunpack::cuda-15%
299.6 µs350.8 µs
iter_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+7%
890.3 ns831.9 ns
set_indices_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+6%
1,039.4 ns981.1 ns
push_n_vortex_buffer[u32, 128]
vortex-buffer/benches/vortex_buffer.rs+5%
1.8 µs1.7 µs
cuda/dict_64vals_bp6bw_codes_for_bp6bw_values/dispatch_u32[100M]
vortex-cuda/benches/dynamic_dispatch_cuda.rs::benches::benchmark_dynamic_dispatch::cuda+5%
1.1 ms1 ms
map_new_output[128]
vortex-buffer/benches/vortex_buffer.rs+4%
760 ns730.8 ns
cuda/alp_f32/0%[100M]
vortex-cuda/benches/alp_cuda.rs::benches::benchmark_alp_decode::cuda+4%
2.5 ms2.4 ms
decompress[u8, (1000, 256)]
encodings/runend/benches/run_end_compress.rs+3%
16 µs15.5 µs
bitwise_not_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
2.1 µs2.1 µs
decompress[u32, (1000, 256)]
encodings/runend/benches/run_end_compress.rs+3%
19 µs18.5 µs
decompress[u8, (1000, 16)]
encodings/runend/benches/run_end_compress.rs+3%
19.2 µs18.7 µs
bitwise_not_arrow_buffer[1024]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
2.3 µs2.2 µs
bitwise_not_arrow_buffer[2048]
vortex-buffer/benches/vortex_bitbuffer.rs+3%
2.4 µs2.3 µs
decompress[u64, (1000, 256)]
encodings/runend/benches/run_end_compress.rs+2%
20.3 µs19.9 µs
decompress[u16, (1000, 256)]
encodings/runend/benches/run_end_compress.rs+2%
18.1 µs17.6 µs
decompress[u32, (1000, 16)]
encodings/runend/benches/run_end_compress.rs+2%
21.9 µs21.4 µs
decompress[u8, (10000, 1024)]
encodings/runend/benches/run_end_compress.rs+2%
22 µs21.5 µs
decompress[u8, (1000, 4)]
encodings/runend/benches/run_end_compress.rs+2%
23.7 µs23.1 µs
decompress[u16, (1000, 16)]
encodings/runend/benches/run_end_compress.rs+2%
20.5 µs20 µs
decompress_utf8[(1000, 256)]
encodings/runend/benches/run_end_compress.rs+2%
26.1 µs25.6 µs
decompress[u64, (1000, 16)]
encodings/runend/benches/run_end_compress.rs+2%
23.3 µs22.8 µs
decompress[u8, (10000, 256)]
encodings/runend/benches/run_end_compress.rs+2%
23.3 µs22.8 µs
append_arrow_buffer[128]
vortex-buffer/benches/vortex_bitbuffer.rs+2%
1.4 µs1.4 µs
decompress_utf8[(1000, 16)]
encodings/runend/benches/run_end_compress.rs+2%
28.2 µs27.7 µs
sum_i32_nulls_clustered
vortex-array/benches/aggregate_sum.rs+2%
240.9 µs236.3 µs
Commits
Click on a commit to change the comparison rangeBase
develop
73454db
-14.6%
test(vortex-turboquant): block-decomposition coverage and docs
962c7fc
4 hours ago
by connortsui20