Finished `bench` profile [optimized] target(s) in 0.03s
Running benchmark/benchmark.rs (target/release/deps/benchmark-1dbf0b03f7f99eab)
╔════════════════════════════════════════════════════════════╗
║ simdna benchmark v1.0.1 ║
║ Run date: 2025-12-18 23:01:19 UTC ║
║ Platform: aarch64 ║
╚════════════════════════════════════════════════════════════╝
Benchmarking encode/simd_4bit/15
Benchmarking encode/simd_4bit/15: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/15: Collecting 100 samples in estimated 5.0000 s (174M iterations)
Benchmarking encode/simd_4bit/15: Analyzing
encode/simd_4bit/15 time: [28.682 ns 28.969 ns 29.353 ns]
thrpt: [487.35 MiB/s 493.81 MiB/s 498.74 MiB/s]
change:
time: [−9.9254% −6.5528% −3.7401%] (p = 0.00 < 0.05)
thrpt: [+3.8854% +7.0123% +11.019%]
Performance has improved.
Found 17 outliers among 100 measurements (17.00%)
8 (8.00%) high mild
9 (9.00%) high severe
Benchmarking encode/scalar_2bit/15
Benchmarking encode/scalar_2bit/15: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/15: Collecting 100 samples in estimated 5.0004 s (53M iterations)
Benchmarking encode/scalar_2bit/15: Analyzing
encode/scalar_2bit/15 time: [94.188 ns 94.264 ns 94.386 ns]
thrpt: [151.56 MiB/s 151.76 MiB/s 151.88 MiB/s]
change:
time: [−7.9751% −5.4861% −3.4814%] (p = 0.00 < 0.05)
thrpt: [+3.6070% +5.8045% +8.6662%]
Performance has improved.
Found 7 outliers among 100 measurements (7.00%)
2 (2.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_4bit/15
Benchmarking encode/scalar_4bit/15: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/15: Collecting 100 samples in estimated 5.0001 s (160M iterations)
Benchmarking encode/scalar_4bit/15: Analyzing
encode/scalar_4bit/15 time: [31.281 ns 31.309 ns 31.350 ns]
thrpt: [456.30 MiB/s 456.89 MiB/s 457.31 MiB/s]
change:
time: [+4.6355% +7.1587% +8.7070%] (p = 0.00 < 0.05)
thrpt: [−8.0096% −6.6804% −4.4302%]
Performance has regressed.
Found 10 outliers among 100 measurements (10.00%)
3 (3.00%) high mild
7 (7.00%) high severe
Benchmarking encode/simd_4bit/16
Benchmarking encode/simd_4bit/16: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/16: Collecting 100 samples in estimated 5.0000 s (186M iterations)
Benchmarking encode/simd_4bit/16: Analyzing
encode/simd_4bit/16 time: [26.802 ns 26.833 ns 26.871 ns]
thrpt: [567.86 MiB/s 568.66 MiB/s 569.31 MiB/s]
change:
time: [−7.6430% −4.8339% −2.7452%] (p = 0.00 < 0.05)
thrpt: [+2.8227% +5.0795% +8.2755%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
5 (5.00%) high mild
3 (3.00%) high severe
Benchmarking encode/scalar_2bit/16
Benchmarking encode/scalar_2bit/16: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/16: Collecting 100 samples in estimated 5.0002 s (99M iterations)
Benchmarking encode/scalar_2bit/16: Analyzing
encode/scalar_2bit/16 time: [50.590 ns 50.622 ns 50.661 ns]
thrpt: [301.19 MiB/s 301.43 MiB/s 301.62 MiB/s]
change:
time: [−7.2551% −5.0751% −3.2373%] (p = 0.00 < 0.05)
thrpt: [+3.3456% +5.3464% +7.8226%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
5 (5.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_4bit/16
Benchmarking encode/scalar_4bit/16: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/16: Collecting 100 samples in estimated 5.0001 s (174M iterations)
Benchmarking encode/scalar_4bit/16: Analyzing
encode/scalar_4bit/16 time: [28.465 ns 28.482 ns 28.502 ns]
thrpt: [535.36 MiB/s 535.73 MiB/s 536.06 MiB/s]
change:
time: [−7.6730% −5.1625% −3.1190%] (p = 0.00 < 0.05)
thrpt: [+3.2194% +5.4435% +8.3107%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
7 (7.00%) high mild
5 (5.00%) high severe
Benchmarking encode/simd_4bit/17
Benchmarking encode/simd_4bit/17: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/17: Collecting 100 samples in estimated 5.0001 s (184M iterations)
Benchmarking encode/simd_4bit/17: Analyzing
encode/simd_4bit/17 time: [26.787 ns 26.804 ns 26.825 ns]
thrpt: [604.37 MiB/s 604.85 MiB/s 605.25 MiB/s]
change:
time: [−6.9235% −4.5472% −2.6912%] (p = 0.00 < 0.05)
thrpt: [+2.7656% +4.7638% +7.4385%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
4 (4.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_2bit/17
Benchmarking encode/scalar_2bit/17: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/17: Collecting 100 samples in estimated 5.0004 s (47M iterations)
Benchmarking encode/scalar_2bit/17: Analyzing
encode/scalar_2bit/17 time: [105.59 ns 105.68 ns 105.79 ns]
thrpt: [153.24 MiB/s 153.42 MiB/s 153.54 MiB/s]
change:
time: [−6.8068% −4.5177% −2.8268%] (p = 0.00 < 0.05)
thrpt: [+2.9090% +4.7315% +7.3039%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
1 (1.00%) low mild
5 (5.00%) high mild
6 (6.00%) high severe
Benchmarking encode/scalar_4bit/17
Benchmarking encode/scalar_4bit/17: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/17: Collecting 100 samples in estimated 5.0001 s (172M iterations)
Benchmarking encode/scalar_4bit/17: Analyzing
encode/scalar_4bit/17 time: [28.987 ns 29.009 ns 29.039 ns]
thrpt: [558.30 MiB/s 558.87 MiB/s 559.30 MiB/s]
change:
time: [−6.5438% −4.5124% −2.9735%] (p = 0.00 < 0.05)
thrpt: [+3.0646% +4.7256% +7.0020%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking encode/simd_4bit/32
Benchmarking encode/simd_4bit/32: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/32: Collecting 100 samples in estimated 5.0001 s (180M iterations)
Benchmarking encode/simd_4bit/32: Analyzing
encode/simd_4bit/32 time: [27.789 ns 27.974 ns 28.316 ns]
thrpt: [1.0525 GiB/s 1.0654 GiB/s 1.0725 GiB/s]
change:
time: [−7.0776% −4.4293% −2.3353%] (p = 0.00 < 0.05)
thrpt: [+2.3912% +4.6345% +7.6167%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
3 (3.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_2bit/32
Benchmarking encode/scalar_2bit/32: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/32: Collecting 100 samples in estimated 5.0003 s (73M iterations)
Benchmarking encode/scalar_2bit/32: Analyzing
encode/scalar_2bit/32 time: [68.103 ns 68.205 ns 68.326 ns]
thrpt: [446.65 MiB/s 447.44 MiB/s 448.11 MiB/s]
change:
time: [−3.8059% −0.8056% +1.5033%] (p = 0.64 > 0.05)
thrpt: [−1.4810% +0.8122% +3.9565%]
No change in performance detected.
Found 6 outliers among 100 measurements (6.00%)
1 (1.00%) low mild
2 (2.00%) high mild
3 (3.00%) high severe
Benchmarking encode/scalar_4bit/32
Benchmarking encode/scalar_4bit/32: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/32: Collecting 100 samples in estimated 5.0001 s (135M iterations)
Benchmarking encode/scalar_4bit/32: Analyzing
encode/scalar_4bit/32 time: [36.964 ns 37.145 ns 37.466 ns]
thrpt: [814.53 MiB/s 821.58 MiB/s 825.61 MiB/s]
change:
time: [−10.132% −7.0515% −4.3110%] (p = 0.00 < 0.05)
thrpt: [+4.5053% +7.5864% +11.275%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
3 (3.00%) high mild
8 (8.00%) high severe
Benchmarking encode/simd_4bit/33
Benchmarking encode/simd_4bit/33: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/33: Collecting 100 samples in estimated 5.0000 s (160M iterations)
Benchmarking encode/simd_4bit/33: Analyzing
encode/simd_4bit/33 time: [31.195 ns 31.291 ns 31.447 ns]
thrpt: [1000.8 MiB/s 1005.7 MiB/s 1008.9 MiB/s]
change:
time: [−8.8303% −5.0691% −1.6947%] (p = 0.00 < 0.05)
thrpt: [+1.7239% +5.3398% +9.6856%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
3 (3.00%) high mild
8 (8.00%) high severe
Benchmarking encode/scalar_2bit/33
Benchmarking encode/scalar_2bit/33: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/33: Collecting 100 samples in estimated 5.0001 s (44M iterations)
Benchmarking encode/scalar_2bit/33: Analyzing
encode/scalar_2bit/33 time: [113.66 ns 113.71 ns 113.77 ns]
thrpt: [276.63 MiB/s 276.77 MiB/s 276.90 MiB/s]
change:
time: [−24.062% −14.980% −6.9311%] (p = 0.00 < 0.05)
thrpt: [+7.4473% +17.620% +31.687%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
2 (2.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_4bit/33
Benchmarking encode/scalar_4bit/33: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/33: Collecting 100 samples in estimated 5.0000 s (123M iterations)
Benchmarking encode/scalar_4bit/33: Analyzing
encode/scalar_4bit/33 time: [40.703 ns 40.756 ns 40.827 ns]
thrpt: [770.83 MiB/s 772.19 MiB/s 773.20 MiB/s]
change:
time: [−1.9727% −1.7306% −1.4383%] (p = 0.00 < 0.05)
thrpt: [+1.4593% +1.7611% +2.0124%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
4 (4.00%) high mild
9 (9.00%) high severe
Benchmarking encode/simd_4bit/63
Benchmarking encode/simd_4bit/63: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/63: Collecting 100 samples in estimated 5.0000 s (123M iterations)
Benchmarking encode/simd_4bit/63: Analyzing
encode/simd_4bit/63 time: [40.570 ns 40.599 ns 40.633 ns]
thrpt: [1.4440 GiB/s 1.4452 GiB/s 1.4462 GiB/s]
change:
time: [−2.7859% −2.6522% −2.5147%] (p = 0.00 < 0.05)
thrpt: [+2.5796% +2.7244% +2.8658%]
Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
2 (2.00%) high mild
4 (4.00%) high severe
Benchmarking encode/scalar_2bit/63
Benchmarking encode/scalar_2bit/63: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/63: Collecting 100 samples in estimated 5.0002 s (39M iterations)
Benchmarking encode/scalar_2bit/63: Analyzing
encode/scalar_2bit/63 time: [129.11 ns 129.30 ns 129.52 ns]
thrpt: [463.88 MiB/s 464.68 MiB/s 465.35 MiB/s]
change:
time: [−3.8814% −3.1020% −2.6180%] (p = 0.00 < 0.05)
thrpt: [+2.6884% +3.2013% +4.0381%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
1 (1.00%) low mild
3 (3.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_4bit/63
Benchmarking encode/scalar_4bit/63: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/63: Collecting 100 samples in estimated 5.0003 s (84M iterations)
Benchmarking encode/scalar_4bit/63: Analyzing
encode/scalar_4bit/63 time: [59.341 ns 59.769 ns 60.404 ns]
thrpt: [994.66 MiB/s 1005.2 MiB/s 1012.5 MiB/s]
change:
time: [−2.2651% −1.4190% −0.4231%] (p = 0.00 < 0.05)
thrpt: [+0.4249% +1.4394% +2.3176%]
Change within noise threshold.
Found 13 outliers among 100 measurements (13.00%)
6 (6.00%) high mild
7 (7.00%) high severe
Benchmarking encode/simd_4bit/64
Benchmarking encode/simd_4bit/64: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/64: Collecting 100 samples in estimated 5.0001 s (133M iterations)
Benchmarking encode/simd_4bit/64: Analyzing
encode/simd_4bit/64 time: [37.465 ns 37.520 ns 37.593 ns]
thrpt: [1.5855 GiB/s 1.5886 GiB/s 1.5910 GiB/s]
change:
time: [−7.5128% −4.7399% −2.5136%] (p = 0.00 < 0.05)
thrpt: [+2.5784% +4.9758% +8.1231%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
5 (5.00%) high mild
9 (9.00%) high severe
Benchmarking encode/scalar_2bit/64
Benchmarking encode/scalar_2bit/64: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/64: Collecting 100 samples in estimated 5.0003 s (60M iterations)
Benchmarking encode/scalar_2bit/64: Analyzing
encode/scalar_2bit/64 time: [82.801 ns 82.855 ns 82.919 ns]
thrpt: [736.08 MiB/s 736.65 MiB/s 737.13 MiB/s]
change:
time: [−9.1266% −6.4501% −4.1788%] (p = 0.00 < 0.05)
thrpt: [+4.3610% +6.8949% +10.043%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
6 (6.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_4bit/64
Benchmarking encode/scalar_4bit/64: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/64: Collecting 100 samples in estimated 5.0002 s (84M iterations)
Benchmarking encode/scalar_4bit/64: Analyzing
encode/scalar_4bit/64 time: [59.701 ns 59.750 ns 59.809 ns]
thrpt: [1020.5 MiB/s 1021.5 MiB/s 1022.4 MiB/s]
change:
time: [−12.686% −8.3319% −4.4741%] (p = 0.00 < 0.05)
thrpt: [+4.6837% +9.0892% +14.529%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
5 (5.00%) high mild
9 (9.00%) high severe
Benchmarking encode/simd_4bit/127
Benchmarking encode/simd_4bit/127: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/127: Collecting 100 samples in estimated 5.0002 s (84M iterations)
Benchmarking encode/simd_4bit/127: Analyzing
encode/simd_4bit/127 time: [59.158 ns 59.194 ns 59.244 ns]
thrpt: [1.9965 GiB/s 1.9981 GiB/s 1.9994 GiB/s]
change:
time: [−7.2269% −4.9502% −3.0501%] (p = 0.00 < 0.05)
thrpt: [+3.1461% +5.2080% +7.7898%]
Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
4 (4.00%) high mild
2 (2.00%) high severe
Benchmarking encode/scalar_2bit/127
Benchmarking encode/scalar_2bit/127: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/127: Collecting 100 samples in estimated 5.0003 s (28M iterations)
Benchmarking encode/scalar_2bit/127: Analyzing
encode/scalar_2bit/127 time: [175.69 ns 175.89 ns 176.14 ns]
thrpt: [687.62 MiB/s 688.58 MiB/s 689.37 MiB/s]
change:
time: [−9.9969% −6.6344% −3.8573%] (p = 0.00 < 0.05)
thrpt: [+4.0120% +7.1059% +11.107%]
Performance has improved.
Found 17 outliers among 100 measurements (17.00%)
3 (3.00%) low mild
7 (7.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_4bit/127
Benchmarking encode/scalar_4bit/127: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/127: Collecting 100 samples in estimated 5.0001 s (45M iterations)
Benchmarking encode/scalar_4bit/127: Analyzing
encode/scalar_4bit/127 time: [112.17 ns 112.48 ns 112.95 ns]
thrpt: [1.0472 GiB/s 1.0515 GiB/s 1.0545 GiB/s]
change:
time: [−9.8301% −5.8489% −2.0777%] (p = 0.00 < 0.05)
thrpt: [+2.1218% +6.2122% +10.902%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking encode/simd_4bit/128
Benchmarking encode/simd_4bit/128: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/128: Collecting 100 samples in estimated 5.0001 s (88M iterations)
Benchmarking encode/simd_4bit/128: Analyzing
encode/simd_4bit/128 time: [56.197 ns 56.239 ns 56.297 ns]
thrpt: [2.1175 GiB/s 2.1197 GiB/s 2.1213 GiB/s]
change:
time: [−8.0066% −4.9757% −2.6265%] (p = 0.00 < 0.05)
thrpt: [+2.6974% +5.2362% +8.7034%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_2bit/128
Benchmarking encode/scalar_2bit/128: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/128: Collecting 100 samples in estimated 5.0005 s (38M iterations)
Benchmarking encode/scalar_2bit/128: Analyzing
encode/scalar_2bit/128 time: [132.54 ns 132.64 ns 132.81 ns]
thrpt: [919.12 MiB/s 920.28 MiB/s 921.04 MiB/s]
change:
time: [−24.253% −17.672% −11.056%] (p = 0.00 < 0.05)
thrpt: [+12.431% +21.465% +32.018%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
1 (1.00%) low mild
3 (3.00%) high mild
4 (4.00%) high severe
Benchmarking encode/scalar_4bit/128
Benchmarking encode/scalar_4bit/128: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/128: Collecting 100 samples in estimated 5.0003 s (44M iterations)
Benchmarking encode/scalar_4bit/128: Analyzing
encode/scalar_4bit/128 time: [112.41 ns 112.51 ns 112.63 ns]
thrpt: [1.0584 GiB/s 1.0596 GiB/s 1.0605 GiB/s]
change:
time: [−8.1299% −5.2212% −2.8951%] (p = 0.00 < 0.05)
thrpt: [+2.9815% +5.5088% +8.8494%]
Performance has improved.
Found 3 outliers among 100 measurements (3.00%)
1 (1.00%) high mild
2 (2.00%) high severe
Benchmarking encode/simd_4bit/255
Benchmarking encode/simd_4bit/255: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/255: Collecting 100 samples in estimated 5.0001 s (55M iterations)
Benchmarking encode/simd_4bit/255: Analyzing
encode/simd_4bit/255 time: [90.928 ns 90.977 ns 91.040 ns]
thrpt: [2.6086 GiB/s 2.6104 GiB/s 2.6118 GiB/s]
change:
time: [−7.0510% −5.0758% −3.5295%] (p = 0.00 < 0.05)
thrpt: [+3.6586% +5.3472% +7.5859%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
7 (7.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_2bit/255
Benchmarking encode/scalar_2bit/255: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/255: Collecting 100 samples in estimated 5.0012 s (20M iterations)
Benchmarking encode/scalar_2bit/255: Analyzing
encode/scalar_2bit/255 time: [255.44 ns 255.58 ns 255.76 ns]
thrpt: [950.84 MiB/s 951.50 MiB/s 952.03 MiB/s]
change:
time: [−5.6748% −3.3821% −1.5942%] (p = 0.00 < 0.05)
thrpt: [+1.6200% +3.5004% +6.0162%]
Performance has improved.
Found 17 outliers among 100 measurements (17.00%)
1 (1.00%) low severe
1 (1.00%) low mild
2 (2.00%) high mild
13 (13.00%) high severe
Benchmarking encode/scalar_4bit/255
Benchmarking encode/scalar_4bit/255: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/255: Collecting 100 samples in estimated 5.0007 s (27M iterations)
Benchmarking encode/scalar_4bit/255: Analyzing
encode/scalar_4bit/255 time: [189.41 ns 191.12 ns 193.98 ns]
thrpt: [1.2243 GiB/s 1.2426 GiB/s 1.2539 GiB/s]
change:
time: [−8.2224% −5.9111% −3.9733%] (p = 0.00 < 0.05)
thrpt: [+4.1377% +6.2824% +8.9590%]
Performance has improved.
Found 4 outliers among 100 measurements (4.00%)
2 (2.00%) high mild
2 (2.00%) high severe
Benchmarking encode/simd_4bit/256
Benchmarking encode/simd_4bit/256: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/256: Collecting 100 samples in estimated 5.0004 s (57M iterations)
Benchmarking encode/simd_4bit/256: Analyzing
encode/simd_4bit/256 time: [88.061 ns 88.670 ns 89.644 ns]
thrpt: [2.6596 GiB/s 2.6888 GiB/s 2.7074 GiB/s]
change:
time: [−3.6293% −3.0406% −2.2066%] (p = 0.00 < 0.05)
thrpt: [+2.2564% +3.1359% +3.7660%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking encode/scalar_2bit/256
Benchmarking encode/scalar_2bit/256: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/256: Collecting 100 samples in estimated 5.0001 s (24M iterations)
Benchmarking encode/scalar_2bit/256: Analyzing
encode/scalar_2bit/256 time: [204.91 ns 205.06 ns 205.24 ns]
thrpt: [1.1617 GiB/s 1.1627 GiB/s 1.1635 GiB/s]
change:
time: [−1.3453% −1.1611% −0.9810%] (p = 0.00 < 0.05)
thrpt: [+0.9907% +1.1748% +1.3636%]
Change within noise threshold.
Found 13 outliers among 100 measurements (13.00%)
4 (4.00%) high mild
9 (9.00%) high severe
Benchmarking encode/scalar_4bit/256
Benchmarking encode/scalar_4bit/256: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/256: Collecting 100 samples in estimated 5.0000 s (26M iterations)
Benchmarking encode/scalar_4bit/256: Analyzing
encode/scalar_4bit/256 time: [188.90 ns 189.03 ns 189.21 ns]
thrpt: [1.2601 GiB/s 1.2613 GiB/s 1.2622 GiB/s]
change:
time: [−4.5887% −4.4056% −4.2322%] (p = 0.00 < 0.05)
thrpt: [+4.4192% +4.6086% +4.8094%]
Performance has improved.
Found 5 outliers among 100 measurements (5.00%)
3 (3.00%) high mild
2 (2.00%) high severe
Benchmarking encode/simd_4bit/512
Benchmarking encode/simd_4bit/512: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/512: Collecting 100 samples in estimated 5.0001 s (32M iterations)
Benchmarking encode/simd_4bit/512: Analyzing
encode/simd_4bit/512 time: [153.53 ns 153.63 ns 153.76 ns]
thrpt: [3.1011 GiB/s 3.1038 GiB/s 3.1058 GiB/s]
change:
time: [−3.2902% −3.1298% −2.9685%] (p = 0.00 < 0.05)
thrpt: [+3.0594% +3.2309% +3.4021%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
4 (4.00%) high mild
9 (9.00%) high severe
Benchmarking encode/scalar_2bit/512
Benchmarking encode/scalar_2bit/512: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/512: Collecting 100 samples in estimated 5.0008 s (14M iterations)
Benchmarking encode/scalar_2bit/512: Analyzing
encode/scalar_2bit/512 time: [359.48 ns 360.54 ns 361.79 ns]
thrpt: [1.3180 GiB/s 1.3226 GiB/s 1.3265 GiB/s]
change:
time: [−8.7252% −5.6871% −3.0136%] (p = 0.00 < 0.05)
thrpt: [+3.1073% +6.0300% +9.5592%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking encode/scalar_4bit/512
Benchmarking encode/scalar_4bit/512: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/512: Collecting 100 samples in estimated 5.0016 s (14M iterations)
Benchmarking encode/scalar_4bit/512: Analyzing
encode/scalar_4bit/512 time: [351.72 ns 352.04 ns 352.41 ns]
thrpt: [1.3531 GiB/s 1.3545 GiB/s 1.3557 GiB/s]
change:
time: [−7.0509% −4.7817% −2.9240%] (p = 0.00 < 0.05)
thrpt: [+3.0120% +5.0218% +7.5857%]
Performance has improved.
Found 3 outliers among 100 measurements (3.00%)
3 (3.00%) high mild
Benchmarking encode/simd_4bit/1023
Benchmarking encode/simd_4bit/1023: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/1023: Collecting 100 samples in estimated 5.0005 s (18M iterations)
Benchmarking encode/simd_4bit/1023: Analyzing
encode/simd_4bit/1023 time: [284.26 ns 287.19 ns 291.57 ns]
thrpt: [3.2676 GiB/s 3.3175 GiB/s 3.3517 GiB/s]
change:
time: [−7.8173% −5.2437% −3.2637%] (p = 0.00 < 0.05)
thrpt: [+3.3738% +5.5339% +8.4802%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
3 (3.00%) high mild
6 (6.00%) high severe
Benchmarking encode/scalar_2bit/1023
Benchmarking encode/scalar_2bit/1023: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/1023: Collecting 100 samples in estimated 5.0020 s (7.0M iterations)
Benchmarking encode/scalar_2bit/1023: Analyzing
encode/scalar_2bit/1023 time: [713.86 ns 714.91 ns 716.20 ns]
thrpt: [1.3303 GiB/s 1.3327 GiB/s 1.3346 GiB/s]
change:
time: [−2.8623% −2.7263% −2.5761%] (p = 0.00 < 0.05)
thrpt: [+2.6443% +2.8027% +2.9466%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
3 (3.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_4bit/1023
Benchmarking encode/scalar_4bit/1023: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/1023: Collecting 100 samples in estimated 5.0021 s (7.4M iterations)
Benchmarking encode/scalar_4bit/1023: Analyzing
encode/scalar_4bit/1023 time: [669.73 ns 670.06 ns 670.45 ns]
thrpt: [1.4210 GiB/s 1.4219 GiB/s 1.4226 GiB/s]
change:
time: [−6.2511% −4.2598% −2.6699%] (p = 0.00 < 0.05)
thrpt: [+2.7432% +4.4494% +6.6679%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
5 (5.00%) high mild
6 (6.00%) high severe
Benchmarking encode/simd_4bit/1024
Benchmarking encode/simd_4bit/1024: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/1024: Collecting 100 samples in estimated 5.0001 s (18M iterations)
Benchmarking encode/simd_4bit/1024: Analyzing
encode/simd_4bit/1024 time: [281.41 ns 283.33 ns 287.21 ns]
thrpt: [3.3204 GiB/s 3.3660 GiB/s 3.3889 GiB/s]
change:
time: [−8.0081% −5.6175% −3.6129%] (p = 0.00 < 0.05)
thrpt: [+3.7483% +5.9518% +8.7052%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
2 (2.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_2bit/1024
Benchmarking encode/scalar_2bit/1024: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/1024: Collecting 100 samples in estimated 5.0006 s (7.6M iterations)
Benchmarking encode/scalar_2bit/1024: Analyzing
encode/scalar_2bit/1024 time: [654.37 ns 655.45 ns 656.76 ns]
thrpt: [1.4521 GiB/s 1.4550 GiB/s 1.4574 GiB/s]
change:
time: [−8.0942% −4.9551% −2.5184%] (p = 0.00 < 0.05)
thrpt: [+2.5834% +5.2134% +8.8071%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
3 (3.00%) high mild
8 (8.00%) high severe
Benchmarking encode/scalar_4bit/1024
Benchmarking encode/scalar_4bit/1024: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/1024: Collecting 100 samples in estimated 5.0015 s (7.4M iterations)
Benchmarking encode/scalar_4bit/1024: Analyzing
encode/scalar_4bit/1024 time: [672.82 ns 673.83 ns 675.27 ns]
thrpt: [1.4123 GiB/s 1.4153 GiB/s 1.4174 GiB/s]
change:
time: [−10.950% −6.9555% −3.2416%] (p = 0.00 < 0.05)
thrpt: [+3.3502% +7.4754% +12.297%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) high mild
12 (12.00%) high severe
Benchmarking encode/simd_4bit/2048
Benchmarking encode/simd_4bit/2048: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/2048: Collecting 100 samples in estimated 5.0023 s (9.1M iterations)
Benchmarking encode/simd_4bit/2048: Analyzing
encode/simd_4bit/2048 time: [545.29 ns 545.68 ns 546.22 ns]
thrpt: [3.4919 GiB/s 3.4953 GiB/s 3.4979 GiB/s]
change:
time: [−10.159% −6.8216% −4.1322%] (p = 0.00 < 0.05)
thrpt: [+4.3103% +7.3210% +11.308%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
3 (3.00%) high mild
6 (6.00%) high severe
Benchmarking encode/scalar_2bit/2048
Benchmarking encode/scalar_2bit/2048: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/2048: Collecting 100 samples in estimated 5.0048 s (4.0M iterations)
Benchmarking encode/scalar_2bit/2048: Analyzing
encode/scalar_2bit/2048 time: [1.2534 µs 1.2572 µs 1.2636 µs]
thrpt: [1.5094 GiB/s 1.5172 GiB/s 1.5218 GiB/s]
change:
time: [−9.1204% −5.6029% −2.7605%] (p = 0.00 < 0.05)
thrpt: [+2.8388% +5.9355% +10.036%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
6 (6.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_4bit/2048
Benchmarking encode/scalar_4bit/2048: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/2048: Collecting 100 samples in estimated 5.0028 s (3.7M iterations)
Benchmarking encode/scalar_4bit/2048: Analyzing
encode/scalar_4bit/2048 time: [1.3348 µs 1.3362 µs 1.3378 µs]
thrpt: [1.4257 GiB/s 1.4275 GiB/s 1.4290 GiB/s]
change:
time: [−2.8649% −2.6898% −2.5144%] (p = 0.00 < 0.05)
thrpt: [+2.5793% +2.7641% +2.9494%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
7 (7.00%) high mild
6 (6.00%) high severe
Benchmarking encode/simd_4bit/4095
Benchmarking encode/simd_4bit/4095: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/4095: Collecting 100 samples in estimated 5.0038 s (4.6M iterations)
Benchmarking encode/simd_4bit/4095: Analyzing
encode/simd_4bit/4095 time: [1.0782 µs 1.0793 µs 1.0808 µs]
thrpt: [3.5287 GiB/s 3.5337 GiB/s 3.5372 GiB/s]
change:
time: [−3.5986% −3.4261% −3.2258%] (p = 0.00 < 0.05)
thrpt: [+3.3333% +3.5477% +3.7329%]
Performance has improved.
Found 16 outliers among 100 measurements (16.00%)
2 (2.00%) low mild
6 (6.00%) high mild
8 (8.00%) high severe
Benchmarking encode/scalar_2bit/4095
Benchmarking encode/scalar_2bit/4095: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/4095: Collecting 100 samples in estimated 5.0062 s (2.0M iterations)
Benchmarking encode/scalar_2bit/4095: Analyzing
encode/scalar_2bit/4095 time: [2.5308 µs 2.5319 µs 2.5333 µs]
thrpt: [1.5055 GiB/s 1.5063 GiB/s 1.5069 GiB/s]
change:
time: [−3.9385% −2.7050% −2.0027%] (p = 0.00 < 0.05)
thrpt: [+2.0436% +2.7802% +4.1000%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
2 (2.00%) high mild
8 (8.00%) high severe
Benchmarking encode/scalar_4bit/4095
Benchmarking encode/scalar_4bit/4095: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/4095: Collecting 100 samples in estimated 5.0032 s (1.9M iterations)
Benchmarking encode/scalar_4bit/4095: Analyzing
encode/scalar_4bit/4095 time: [2.6100 µs 2.6244 µs 2.6517 µs]
thrpt: [1.4382 GiB/s 1.4532 GiB/s 1.4612 GiB/s]
change:
time: [−6.4035% −4.3325% −2.6874%] (p = 0.00 < 0.05)
thrpt: [+2.7616% +4.5287% +6.8416%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) low mild
3 (3.00%) high mild
9 (9.00%) high severe
Benchmarking encode/simd_4bit/4096
Benchmarking encode/simd_4bit/4096: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/4096: Collecting 100 samples in estimated 5.0037 s (4.6M iterations)
Benchmarking encode/simd_4bit/4096: Analyzing
encode/simd_4bit/4096 time: [1.0843 µs 1.0891 µs 1.0986 µs]
thrpt: [3.4723 GiB/s 3.5025 GiB/s 3.5180 GiB/s]
change:
time: [−6.0015% −4.0177% −2.4503%] (p = 0.00 < 0.05)
thrpt: [+2.5119% +4.1859% +6.3846%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
4 (4.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_2bit/4096
Benchmarking encode/scalar_2bit/4096: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/4096: Collecting 100 samples in estimated 5.0020 s (2.1M iterations)
Benchmarking encode/scalar_2bit/4096: Analyzing
encode/scalar_2bit/4096 time: [2.4253 µs 2.4270 µs 2.4290 µs]
thrpt: [1.5705 GiB/s 1.5718 GiB/s 1.5729 GiB/s]
change:
time: [−7.8110% −5.0725% −2.8280%] (p = 0.00 < 0.05)
thrpt: [+2.9103% +5.3436% +8.4728%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
1 (1.00%) low mild
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_4bit/4096
Benchmarking encode/scalar_4bit/4096: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/4096: Collecting 100 samples in estimated 5.0011 s (1.9M iterations)
Benchmarking encode/scalar_4bit/4096: Analyzing
encode/scalar_4bit/4096 time: [2.6103 µs 2.6121 µs 2.6143 µs]
thrpt: [1.4592 GiB/s 1.4604 GiB/s 1.4614 GiB/s]
change:
time: [−2.6323% −2.4620% −2.3070%] (p = 0.00 < 0.05)
thrpt: [+2.3615% +2.5242% +2.7034%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
6 (6.00%) high mild
6 (6.00%) high severe
Benchmarking encode/simd_4bit/8192
Benchmarking encode/simd_4bit/8192: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/8192: Collecting 100 samples in estimated 5.0008 s (2.3M iterations)
Benchmarking encode/simd_4bit/8192: Analyzing
encode/simd_4bit/8192 time: [2.1416 µs 2.1434 µs 2.1459 µs]
thrpt: [3.5553 GiB/s 3.5595 GiB/s 3.5626 GiB/s]
change:
time: [−2.5584% −2.3565% −2.1368%] (p = 0.00 < 0.05)
thrpt: [+2.1835% +2.4134% +2.6256%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
2 (2.00%) high mild
8 (8.00%) high severe
Benchmarking encode/scalar_2bit/8192
Benchmarking encode/scalar_2bit/8192: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/8192: Collecting 100 samples in estimated 5.0125 s (1.1M iterations)
Benchmarking encode/scalar_2bit/8192: Analyzing
encode/scalar_2bit/8192 time: [4.7642 µs 4.8092 µs 4.8753 µs]
thrpt: [1.5649 GiB/s 1.5864 GiB/s 1.6014 GiB/s]
change:
time: [−2.6914% −2.1702% −1.4550%] (p = 0.00 < 0.05)
thrpt: [+1.4765% +2.2184% +2.7658%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
6 (6.00%) high mild
9 (9.00%) high severe
Benchmarking encode/scalar_4bit/8192
Benchmarking encode/scalar_4bit/8192: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/8192: Collecting 100 samples in estimated 5.0219 s (970k iterations)
Benchmarking encode/scalar_4bit/8192: Analyzing
encode/scalar_4bit/8192 time: [5.1796 µs 5.2087 µs 5.2548 µs]
thrpt: [1.4519 GiB/s 1.4647 GiB/s 1.4730 GiB/s]
change:
time: [−5.6229% −3.5506% −1.9904%] (p = 0.00 < 0.05)
thrpt: [+2.0308% +3.6813% +5.9579%]
Performance has improved.
Found 19 outliers among 100 measurements (19.00%)
1 (1.00%) low mild
8 (8.00%) high mild
10 (10.00%) high severe
Benchmarking encode/simd_4bit/9999
Benchmarking encode/simd_4bit/9999: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/9999: Collecting 100 samples in estimated 5.0059 s (1.9M iterations)
Benchmarking encode/simd_4bit/9999: Analyzing
encode/simd_4bit/9999 time: [2.6205 µs 2.6227 µs 2.6255 µs]
thrpt: [3.5469 GiB/s 3.5507 GiB/s 3.5537 GiB/s]
change:
time: [−6.5716% −4.3890% −2.7087%] (p = 0.00 < 0.05)
thrpt: [+2.7841% +4.5905% +7.0339%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
7 (7.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_2bit/9999
Benchmarking encode/scalar_2bit/9999: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/9999: Collecting 100 samples in estimated 5.0103 s (838k iterations)
Benchmarking encode/scalar_2bit/9999: Analyzing
encode/scalar_2bit/9999 time: [5.9676 µs 5.9715 µs 5.9766 µs]
thrpt: [1.5581 GiB/s 1.5595 GiB/s 1.5605 GiB/s]
change:
time: [−7.2158% −4.7419% −2.8093%] (p = 0.00 < 0.05)
thrpt: [+2.8905% +4.9779% +7.7769%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
3 (3.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_4bit/9999
Benchmarking encode/scalar_4bit/9999: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/9999: Collecting 100 samples in estimated 5.0239 s (793k iterations)
Benchmarking encode/scalar_4bit/9999: Analyzing
encode/scalar_4bit/9999 time: [6.3284 µs 6.3322 µs 6.3370 µs]
thrpt: [1.4695 GiB/s 1.4706 GiB/s 1.4715 GiB/s]
change:
time: [−6.8919% −4.3845% −2.5593%] (p = 0.00 < 0.05)
thrpt: [+2.6265% +4.5856% +7.4021%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
3 (3.00%) high mild
5 (5.00%) high severe
Benchmarking encode/simd_4bit/10000
Benchmarking encode/simd_4bit/10000: Warming up for 3.0000 s
Benchmarking encode/simd_4bit/10000: Collecting 100 samples in estimated 5.0117 s (1.9M iterations)
Benchmarking encode/simd_4bit/10000: Analyzing
encode/simd_4bit/10000 time: [2.6158 µs 2.6176 µs 2.6201 µs]
thrpt: [3.5545 GiB/s 3.5579 GiB/s 3.5603 GiB/s]
change:
time: [−7.1777% −5.0450% −3.2852%] (p = 0.00 < 0.05)
thrpt: [+3.3968% +5.3130% +7.7327%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
2 (2.00%) high mild
7 (7.00%) high severe
Benchmarking encode/scalar_2bit/10000
Benchmarking encode/scalar_2bit/10000: Warming up for 3.0000 s
Benchmarking encode/scalar_2bit/10000: Collecting 100 samples in estimated 5.0144 s (864k iterations)
Benchmarking encode/scalar_2bit/10000: Analyzing
encode/scalar_2bit/10000
time: [5.7885 µs 5.7923 µs 5.7978 µs]
thrpt: [1.6063 GiB/s 1.6079 GiB/s 1.6089 GiB/s]
change:
time: [−7.9622% −5.2748% −3.1425%] (p = 0.00 < 0.05)
thrpt: [+3.2445% +5.5685% +8.6510%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) low mild
7 (7.00%) high mild
5 (5.00%) high severe
Benchmarking encode/scalar_4bit/10000
Benchmarking encode/scalar_4bit/10000: Warming up for 3.0000 s
Benchmarking encode/scalar_4bit/10000: Collecting 100 samples in estimated 5.0094 s (783k iterations)
Benchmarking encode/scalar_4bit/10000: Analyzing
encode/scalar_4bit/10000
time: [6.3252 µs 6.3456 µs 6.3818 µs]
thrpt: [1.4593 GiB/s 1.4677 GiB/s 1.4724 GiB/s]
change:
time: [−5.7173% −3.9249% −2.4827%] (p = 0.00 < 0.05)
thrpt: [+2.5459% +4.0852% +6.0640%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
2 (2.00%) low mild
2 (2.00%) high mild
8 (8.00%) high severe
Benchmarking decode/simd_4bit/15
Benchmarking decode/simd_4bit/15: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/15: Collecting 100 samples in estimated 5.0000 s (171M iterations)
Benchmarking decode/simd_4bit/15: Analyzing
decode/simd_4bit/15 time: [29.147 ns 29.162 ns 29.180 ns]
thrpt: [490.24 MiB/s 490.53 MiB/s 490.79 MiB/s]
change:
time: [−7.5753% −4.9710% −2.9325%] (p = 0.00 < 0.05)
thrpt: [+3.0211% +5.2310% +8.1961%]
Performance has improved.
Found 16 outliers among 100 measurements (16.00%)
1 (1.00%) low mild
5 (5.00%) high mild
10 (10.00%) high severe
Benchmarking decode/scalar_2bit/15
Benchmarking decode/scalar_2bit/15: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/15: Collecting 100 samples in estimated 5.0001 s (171M iterations)
Benchmarking decode/scalar_2bit/15: Analyzing
decode/scalar_2bit/15 time: [29.194 ns 29.208 ns 29.223 ns]
thrpt: [489.52 MiB/s 489.77 MiB/s 489.99 MiB/s]
change:
time: [+4.5789% +7.2819% +9.4721%] (p = 0.00 < 0.05)
thrpt: [−8.6525% −6.7876% −4.3784%]
Performance has regressed.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking decode/scalar_4bit/15
Benchmarking decode/scalar_4bit/15: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/15: Collecting 100 samples in estimated 5.0001 s (201M iterations)
Benchmarking decode/scalar_4bit/15: Analyzing
decode/scalar_4bit/15 time: [25.080 ns 25.221 ns 25.376 ns]
thrpt: [563.73 MiB/s 567.19 MiB/s 570.37 MiB/s]
change:
time: [−3.2316% −2.5250% −1.8087%] (p = 0.00 < 0.05)
thrpt: [+1.8420% +2.5904% +3.3395%]
Performance has improved.
Benchmarking decode/simd_4bit/16
Benchmarking decode/simd_4bit/16: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/16: Collecting 100 samples in estimated 5.0001 s (187M iterations)
Benchmarking decode/simd_4bit/16: Analyzing
decode/simd_4bit/16 time: [26.733 ns 26.753 ns 26.779 ns]
thrpt: [569.81 MiB/s 570.37 MiB/s 570.78 MiB/s]
change:
time: [−3.5704% −3.3565% −3.1494%] (p = 0.00 < 0.05)
thrpt: [+3.2518% +3.4731% +3.7026%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
1 (1.00%) low mild
6 (6.00%) high mild
3 (3.00%) high severe
Benchmarking decode/scalar_2bit/16
Benchmarking decode/scalar_2bit/16: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/16: Collecting 100 samples in estimated 5.0001 s (170M iterations)
Benchmarking decode/scalar_2bit/16: Analyzing
decode/scalar_2bit/16 time: [29.450 ns 29.487 ns 29.532 ns]
thrpt: [516.69 MiB/s 517.47 MiB/s 518.12 MiB/s]
change:
time: [+9.7767% +9.9978% +10.252%] (p = 0.00 < 0.05)
thrpt: [−9.2983% −9.0891% −8.9060%]
Performance has regressed.
Found 12 outliers among 100 measurements (12.00%)
1 (1.00%) high mild
11 (11.00%) high severe
Benchmarking decode/scalar_4bit/16
Benchmarking decode/scalar_4bit/16: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/16: Collecting 100 samples in estimated 5.0000 s (199M iterations)
Benchmarking decode/scalar_4bit/16: Analyzing
decode/scalar_4bit/16 time: [25.333 ns 25.474 ns 25.649 ns]
thrpt: [594.90 MiB/s 598.99 MiB/s 602.33 MiB/s]
change:
time: [−2.5723% −1.8775% −1.0821%] (p = 0.00 < 0.05)
thrpt: [+1.0940% +1.9134% +2.6402%]
Performance has improved.
Found 2 outliers among 100 measurements (2.00%)
1 (1.00%) high mild
1 (1.00%) high severe
Benchmarking decode/simd_4bit/17
Benchmarking decode/simd_4bit/17: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/17: Collecting 100 samples in estimated 5.0000 s (187M iterations)
Benchmarking decode/simd_4bit/17: Analyzing
decode/simd_4bit/17 time: [26.784 ns 26.841 ns 26.895 ns]
thrpt: [602.81 MiB/s 604.01 MiB/s 605.30 MiB/s]
change:
time: [−13.806% −8.9475% −4.3737%] (p = 0.00 < 0.05)
thrpt: [+4.5738% +9.8268% +16.017%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
1 (1.00%) low mild
5 (5.00%) high mild
2 (2.00%) high severe
Benchmarking decode/scalar_2bit/17
Benchmarking decode/scalar_2bit/17: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/17: Collecting 100 samples in estimated 5.0001 s (170M iterations)
Benchmarking decode/scalar_2bit/17: Analyzing
decode/scalar_2bit/17 time: [28.945 ns 28.969 ns 28.997 ns]
thrpt: [559.11 MiB/s 559.65 MiB/s 560.12 MiB/s]
change:
time: [−9.3230% −6.2537% −3.6847%] (p = 0.00 < 0.05)
thrpt: [+3.8257% +6.6709% +10.281%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
4 (4.00%) high mild
9 (9.00%) high severe
Benchmarking decode/scalar_4bit/17
Benchmarking decode/scalar_4bit/17: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/17: Collecting 100 samples in estimated 5.0001 s (182M iterations)
Benchmarking decode/scalar_4bit/17: Analyzing
decode/scalar_4bit/17 time: [27.451 ns 27.478 ns 27.511 ns]
thrpt: [589.31 MiB/s 590.01 MiB/s 590.59 MiB/s]
change:
time: [−10.002% −7.0692% −4.5919%] (p = 0.00 < 0.05)
thrpt: [+4.8130% +7.6069% +11.114%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
8 (8.00%) high mild
7 (7.00%) high severe
Benchmarking decode/simd_4bit/32
Benchmarking decode/simd_4bit/32: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/32: Collecting 100 samples in estimated 5.0001 s (176M iterations)
Benchmarking decode/simd_4bit/32: Analyzing
decode/simd_4bit/32 time: [28.437 ns 28.474 ns 28.520 ns]
thrpt: [1.0450 GiB/s 1.0466 GiB/s 1.0480 GiB/s]
change:
time: [−10.773% −6.6079% −3.0140%] (p = 0.00 < 0.05)
thrpt: [+3.1076% +7.0754% +12.074%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
5 (5.00%) high mild
7 (7.00%) high severe
Benchmarking decode/scalar_2bit/32
Benchmarking decode/scalar_2bit/32: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/32: Collecting 100 samples in estimated 5.0001 s (147M iterations)
Benchmarking decode/scalar_2bit/32: Analyzing
decode/scalar_2bit/32 time: [33.667 ns 33.709 ns 33.762 ns]
thrpt: [903.89 MiB/s 905.32 MiB/s 906.45 MiB/s]
change:
time: [−15.847% −11.312% −7.2120%] (p = 0.00 < 0.05)
thrpt: [+7.7726% +12.755% +18.832%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
5 (5.00%) high mild
9 (9.00%) high severe
Benchmarking decode/scalar_4bit/32
Benchmarking decode/scalar_4bit/32: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/32: Collecting 100 samples in estimated 5.0000 s (148M iterations)
Benchmarking decode/scalar_4bit/32: Analyzing
decode/scalar_4bit/32 time: [33.594 ns 33.821 ns 34.232 ns]
thrpt: [891.50 MiB/s 902.33 MiB/s 908.42 MiB/s]
change:
time: [−18.089% −11.396% −5.8256%] (p = 0.00 < 0.05)
thrpt: [+6.1859% +12.861% +22.083%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
1 (1.00%) high mild
10 (10.00%) high severe
Benchmarking decode/simd_4bit/33
Benchmarking decode/simd_4bit/33: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/33: Collecting 100 samples in estimated 5.0001 s (172M iterations)
Benchmarking decode/simd_4bit/33: Analyzing
decode/simd_4bit/33 time: [29.167 ns 29.396 ns 29.750 ns]
thrpt: [1.0331 GiB/s 1.0455 GiB/s 1.0537 GiB/s]
change:
time: [−3.3002% −2.7595% −2.1612%] (p = 0.00 < 0.05)
thrpt: [+2.2089% +2.8379% +3.4128%]
Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
2 (2.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_2bit/33
Benchmarking decode/scalar_2bit/33: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/33: Collecting 100 samples in estimated 5.0002 s (146M iterations)
Benchmarking decode/scalar_2bit/33: Analyzing
decode/scalar_2bit/33 time: [34.319 ns 34.341 ns 34.375 ns]
thrpt: [915.54 MiB/s 916.43 MiB/s 917.03 MiB/s]
change:
time: [−9.7619% −9.6234% −9.4851%] (p = 0.00 < 0.05)
thrpt: [+10.479% +10.648% +10.818%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
1 (1.00%) low mild
2 (2.00%) high mild
7 (7.00%) high severe
Benchmarking decode/scalar_4bit/33
Benchmarking decode/scalar_4bit/33: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/33: Collecting 100 samples in estimated 5.0001 s (144M iterations)
Benchmarking decode/scalar_4bit/33: Analyzing
decode/scalar_4bit/33 time: [34.624 ns 34.645 ns 34.674 ns]
thrpt: [907.64 MiB/s 908.38 MiB/s 908.94 MiB/s]
change:
time: [−1.9972% −1.8279% −1.6617%] (p = 0.00 < 0.05)
thrpt: [+1.6897% +1.8620% +2.0379%]
Performance has improved.
Found 16 outliers among 100 measurements (16.00%)
1 (1.00%) low mild
7 (7.00%) high mild
8 (8.00%) high severe
Benchmarking decode/simd_4bit/63
Benchmarking decode/simd_4bit/63: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/63: Collecting 100 samples in estimated 5.0001 s (146M iterations)
Benchmarking decode/simd_4bit/63: Analyzing
decode/simd_4bit/63 time: [34.253 ns 34.282 ns 34.318 ns]
thrpt: [1.7097 GiB/s 1.7115 GiB/s 1.7130 GiB/s]
change:
time: [−10.486% −10.341% −10.191%] (p = 0.00 < 0.05)
thrpt: [+11.347% +11.534% +11.714%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_2bit/63
Benchmarking decode/scalar_2bit/63: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/63: Collecting 100 samples in estimated 5.0002 s (119M iterations)
Benchmarking decode/scalar_2bit/63: Analyzing
decode/scalar_2bit/63 time: [41.998 ns 42.025 ns 42.057 ns]
thrpt: [1.3951 GiB/s 1.3962 GiB/s 1.3970 GiB/s]
change:
time: [−3.0702% −2.3140% −1.8482%] (p = 0.00 < 0.05)
thrpt: [+1.8830% +2.3688% +3.1675%]
Performance has improved.
Found 16 outliers among 100 measurements (16.00%)
6 (6.00%) high mild
10 (10.00%) high severe
Benchmarking decode/scalar_4bit/63
Benchmarking decode/scalar_4bit/63: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/63: Collecting 100 samples in estimated 5.0000 s (113M iterations)
Benchmarking decode/scalar_4bit/63: Analyzing
decode/scalar_4bit/63 time: [44.217 ns 44.517 ns 44.981 ns]
thrpt: [1.3044 GiB/s 1.3180 GiB/s 1.3269 GiB/s]
change:
time: [−8.6881% −5.6312% −2.9314%] (p = 0.00 < 0.05)
thrpt: [+3.0199% +5.9673% +9.5148%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking decode/simd_4bit/64
Benchmarking decode/simd_4bit/64: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/64: Collecting 100 samples in estimated 5.0000 s (156M iterations)
Benchmarking decode/simd_4bit/64: Analyzing
decode/simd_4bit/64 time: [32.129 ns 32.282 ns 32.518 ns]
thrpt: [1.8330 GiB/s 1.8464 GiB/s 1.8552 GiB/s]
change:
time: [−20.338% −15.585% −11.547%] (p = 0.00 < 0.05)
thrpt: [+13.055% +18.462% +25.530%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
7 (7.00%) high mild
7 (7.00%) high severe
Benchmarking decode/scalar_2bit/64
Benchmarking decode/scalar_2bit/64: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/64: Collecting 100 samples in estimated 5.0001 s (118M iterations)
Benchmarking decode/scalar_2bit/64: Analyzing
decode/scalar_2bit/64 time: [42.375 ns 42.391 ns 42.409 ns]
thrpt: [1.4055 GiB/s 1.4061 GiB/s 1.4066 GiB/s]
change:
time: [−2.3426% −2.1553% −1.9450%] (p = 0.00 < 0.05)
thrpt: [+1.9836% +2.2028% +2.3988%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking decode/scalar_4bit/64
Benchmarking decode/scalar_4bit/64: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/64: Collecting 100 samples in estimated 5.0002 s (113M iterations)
Benchmarking decode/scalar_4bit/64: Analyzing
decode/scalar_4bit/64 time: [44.338 ns 44.374 ns 44.429 ns]
thrpt: [1.3416 GiB/s 1.3432 GiB/s 1.3443 GiB/s]
change:
time: [+1.7972% +2.0603% +2.3255%] (p = 0.00 < 0.05)
thrpt: [−2.2726% −2.0187% −1.7655%]
Performance has regressed.
Found 9 outliers among 100 measurements (9.00%)
1 (1.00%) low mild
1 (1.00%) high mild
7 (7.00%) high severe
Benchmarking decode/simd_4bit/127
Benchmarking decode/simd_4bit/127: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/127: Collecting 100 samples in estimated 5.0002 s (103M iterations)
Benchmarking decode/simd_4bit/127: Analyzing
decode/simd_4bit/127 time: [48.255 ns 48.305 ns 48.375 ns]
thrpt: [2.4450 GiB/s 2.4486 GiB/s 2.4511 GiB/s]
change:
time: [−22.120% −14.817% −7.8581%] (p = 0.00 < 0.05)
thrpt: [+8.5283% +17.394% +28.402%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_2bit/127
Benchmarking decode/scalar_2bit/127: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/127: Collecting 100 samples in estimated 5.0003 s (75M iterations)
Benchmarking decode/scalar_2bit/127: Analyzing
decode/scalar_2bit/127 time: [66.622 ns 66.706 ns 66.816 ns]
thrpt: [1.7702 GiB/s 1.7731 GiB/s 1.7754 GiB/s]
change:
time: [+2.1427% +2.7097% +3.1385%] (p = 0.00 < 0.05)
thrpt: [−3.0430% −2.6383% −2.0978%]
Performance has regressed.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) high mild
12 (12.00%) high severe
Benchmarking decode/scalar_4bit/127
Benchmarking decode/scalar_4bit/127: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/127: Collecting 100 samples in estimated 5.0001 s (81M iterations)
Benchmarking decode/scalar_4bit/127: Analyzing
decode/scalar_4bit/127 time: [61.489 ns 61.711 ns 62.112 ns]
thrpt: [1.9043 GiB/s 1.9167 GiB/s 1.9236 GiB/s]
change:
time: [−4.3466% −3.4892% −2.5309%] (p = 0.00 < 0.05)
thrpt: [+2.5966% +3.6154% +4.5441%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
3 (3.00%) high mild
11 (11.00%) high severe
Benchmarking decode/simd_4bit/128
Benchmarking decode/simd_4bit/128: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/128: Collecting 100 samples in estimated 5.0002 s (104M iterations)
Benchmarking decode/simd_4bit/128: Analyzing
decode/simd_4bit/128 time: [47.814 ns 48.002 ns 48.304 ns]
thrpt: [2.4679 GiB/s 2.4834 GiB/s 2.4932 GiB/s]
change:
time: [−2.4237% −1.8759% −1.1004%] (p = 0.00 < 0.05)
thrpt: [+1.1127% +1.9118% +2.4839%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
4 (4.00%) high mild
10 (10.00%) high severe
Benchmarking decode/scalar_2bit/128
Benchmarking decode/scalar_2bit/128: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/128: Collecting 100 samples in estimated 5.0002 s (75M iterations)
Benchmarking decode/scalar_2bit/128: Analyzing
decode/scalar_2bit/128 time: [66.916 ns 66.956 ns 67.007 ns]
thrpt: [1.7790 GiB/s 1.7804 GiB/s 1.7815 GiB/s]
change:
time: [+2.4146% +2.5932% +2.7764%] (p = 0.00 < 0.05)
thrpt: [−2.7014% −2.5277% −2.3577%]
Performance has regressed.
Found 16 outliers among 100 measurements (16.00%)
2 (2.00%) low mild
6 (6.00%) high mild
8 (8.00%) high severe
Benchmarking decode/scalar_4bit/128
Benchmarking decode/scalar_4bit/128: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/128: Collecting 100 samples in estimated 5.0001 s (81M iterations)
Benchmarking decode/scalar_4bit/128: Analyzing
decode/scalar_4bit/128 time: [61.800 ns 61.835 ns 61.880 ns]
thrpt: [1.9265 GiB/s 1.9279 GiB/s 1.9290 GiB/s]
change:
time: [−3.3936% −3.2132% −3.0453%] (p = 0.00 < 0.05)
thrpt: [+3.1409% +3.3198% +3.5128%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
4 (4.00%) high mild
7 (7.00%) high severe
Benchmarking decode/simd_4bit/255
Benchmarking decode/simd_4bit/255: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/255: Collecting 100 samples in estimated 5.0003 s (71M iterations)
Benchmarking decode/simd_4bit/255: Analyzing
decode/simd_4bit/255 time: [70.420 ns 70.462 ns 70.515 ns]
thrpt: [3.3679 GiB/s 3.3704 GiB/s 3.3724 GiB/s]
change:
time: [−6.6812% −4.4354% −2.7512%] (p = 0.00 < 0.05)
thrpt: [+2.8291% +4.6412% +7.1596%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_2bit/255
Benchmarking decode/scalar_2bit/255: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/255: Collecting 100 samples in estimated 5.0005 s (45M iterations)
Benchmarking decode/scalar_2bit/255: Analyzing
decode/scalar_2bit/255 time: [110.42 ns 110.50 ns 110.61 ns]
thrpt: [2.1471 GiB/s 2.1492 GiB/s 2.1507 GiB/s]
change:
time: [−7.3716% −4.5111% −2.4213%] (p = 0.00 < 0.05)
thrpt: [+2.4814% +4.7242% +7.9583%]
Performance has improved.
Found 17 outliers among 100 measurements (17.00%)
1 (1.00%) low mild
6 (6.00%) high mild
10 (10.00%) high severe
Benchmarking decode/scalar_4bit/255
Benchmarking decode/scalar_4bit/255: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/255: Collecting 100 samples in estimated 5.0006 s (43M iterations)
Benchmarking decode/scalar_4bit/255: Analyzing
decode/scalar_4bit/255 time: [115.34 ns 116.49 ns 118.22 ns]
thrpt: [2.0089 GiB/s 2.0387 GiB/s 2.0590 GiB/s]
change:
time: [−5.7105% −2.6932% −0.2897%] (p = 0.04 < 0.05)
thrpt: [+0.2906% +2.7677% +6.0564%]
Change within noise threshold.
Found 8 outliers among 100 measurements (8.00%)
5 (5.00%) high mild
3 (3.00%) high severe
Benchmarking decode/simd_4bit/256
Benchmarking decode/simd_4bit/256: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/256: Collecting 100 samples in estimated 5.0001 s (73M iterations)
Benchmarking decode/simd_4bit/256: Analyzing
decode/simd_4bit/256 time: [68.529 ns 69.084 ns 69.851 ns]
thrpt: [3.4133 GiB/s 3.4512 GiB/s 3.4791 GiB/s]
change:
time: [−12.305% −7.3128% −3.3605%] (p = 0.00 < 0.05)
thrpt: [+3.4774% +7.8898% +14.032%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
2 (2.00%) high mild
7 (7.00%) high severe
Benchmarking decode/scalar_2bit/256
Benchmarking decode/scalar_2bit/256: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/256: Collecting 100 samples in estimated 5.0003 s (45M iterations)
Benchmarking decode/scalar_2bit/256: Analyzing
decode/scalar_2bit/256 time: [110.92 ns 111.01 ns 111.11 ns]
thrpt: [2.1458 GiB/s 2.1478 GiB/s 2.1494 GiB/s]
change:
time: [−7.4010% −4.9252% −2.9358%] (p = 0.00 < 0.05)
thrpt: [+3.0246% +5.1804% +7.9925%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
1 (1.00%) low mild
3 (3.00%) high mild
5 (5.00%) high severe
Benchmarking decode/scalar_4bit/256
Benchmarking decode/scalar_4bit/256: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/256: Collecting 100 samples in estimated 5.0002 s (44M iterations)
Benchmarking decode/scalar_4bit/256: Analyzing
decode/scalar_4bit/256 time: [112.30 ns 112.39 ns 112.53 ns]
thrpt: [2.1187 GiB/s 2.1213 GiB/s 2.1231 GiB/s]
change:
time: [−4.4918% −2.4908% −1.0533%] (p = 0.00 < 0.05)
thrpt: [+1.0645% +2.5544% +4.7030%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
2 (2.00%) low mild
8 (8.00%) high mild
5 (5.00%) high severe
Benchmarking decode/simd_4bit/512
Benchmarking decode/simd_4bit/512: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/512: Collecting 100 samples in estimated 5.0005 s (46M iterations)
Benchmarking decode/simd_4bit/512: Analyzing
decode/simd_4bit/512 time: [108.37 ns 108.42 ns 108.47 ns]
thrpt: [4.3961 GiB/s 4.3981 GiB/s 4.3999 GiB/s]
change:
time: [−12.097% −8.0578% −4.6783%] (p = 0.00 < 0.05)
thrpt: [+4.9079% +8.7640% +13.762%]
Performance has improved.
Found 7 outliers among 100 measurements (7.00%)
3 (3.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_2bit/512
Benchmarking decode/scalar_2bit/512: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/512: Collecting 100 samples in estimated 5.0003 s (25M iterations)
Benchmarking decode/scalar_2bit/512: Analyzing
decode/scalar_2bit/512 time: [202.50 ns 202.77 ns 203.14 ns]
thrpt: [2.3474 GiB/s 2.3516 GiB/s 2.3548 GiB/s]
change:
time: [−4.5225% −2.9690% −1.7322%] (p = 0.00 < 0.05)
thrpt: [+1.7627% +3.0599% +4.7367%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
9 (9.00%) high mild
2 (2.00%) high severe
Benchmarking decode/scalar_4bit/512
Benchmarking decode/scalar_4bit/512: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/512: Collecting 100 samples in estimated 5.0004 s (26M iterations)
Benchmarking decode/scalar_4bit/512: Analyzing
decode/scalar_4bit/512 time: [190.73 ns 190.93 ns 191.19 ns]
thrpt: [2.4940 GiB/s 2.4974 GiB/s 2.5001 GiB/s]
change:
time: [−6.9158% −4.2518% −2.1372%] (p = 0.00 < 0.05)
thrpt: [+2.1839% +4.4407% +7.4297%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
5 (5.00%) high mild
7 (7.00%) high severe
Benchmarking decode/simd_4bit/1023
Benchmarking decode/simd_4bit/1023: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/1023: Collecting 100 samples in estimated 5.0005 s (25M iterations)
Benchmarking decode/simd_4bit/1023: Analyzing
decode/simd_4bit/1023 time: [196.52 ns 197.09 ns 197.97 ns]
thrpt: [4.8126 GiB/s 4.8340 GiB/s 4.8482 GiB/s]
change:
time: [−7.3024% −3.7124% −0.2308%] (p = 0.04 < 0.05)
thrpt: [+0.2313% +3.8555% +7.8776%]
Change within noise threshold.
Found 13 outliers among 100 measurements (13.00%)
6 (6.00%) high mild
7 (7.00%) high severe
Benchmarking decode/scalar_2bit/1023
Benchmarking decode/scalar_2bit/1023: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/1023: Collecting 100 samples in estimated 5.0017 s (13M iterations)
Benchmarking decode/scalar_2bit/1023: Analyzing
decode/scalar_2bit/1023 time: [387.48 ns 388.19 ns 389.55 ns]
thrpt: [2.4458 GiB/s 2.4543 GiB/s 2.4588 GiB/s]
change:
time: [−6.4623% −3.7651% −1.1562%] (p = 0.00 < 0.05)
thrpt: [+1.1697% +3.9124% +6.9087%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
1 (1.00%) low mild
5 (5.00%) high mild
5 (5.00%) high severe
Benchmarking decode/scalar_4bit/1023
Benchmarking decode/scalar_4bit/1023: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/1023: Collecting 100 samples in estimated 5.0011 s (14M iterations)
Benchmarking decode/scalar_4bit/1023: Analyzing
decode/scalar_4bit/1023 time: [355.18 ns 355.50 ns 355.88 ns]
thrpt: [2.6771 GiB/s 2.6800 GiB/s 2.6824 GiB/s]
change:
time: [−8.9032% −6.0167% −3.6475%] (p = 0.00 < 0.05)
thrpt: [+3.7856% +6.4018% +9.7734%]
Performance has improved.
Found 5 outliers among 100 measurements (5.00%)
3 (3.00%) high mild
2 (2.00%) high severe
Benchmarking decode/simd_4bit/1024
Benchmarking decode/simd_4bit/1024: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/1024: Collecting 100 samples in estimated 5.0001 s (26M iterations)
Benchmarking decode/simd_4bit/1024: Analyzing
decode/simd_4bit/1024 time: [195.01 ns 195.08 ns 195.17 ns]
thrpt: [4.8864 GiB/s 4.8886 GiB/s 4.8905 GiB/s]
change:
time: [−8.0570% −4.8149% −2.2900%] (p = 0.00 < 0.05)
thrpt: [+2.3437% +5.0585% +8.7631%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
2 (2.00%) high mild
7 (7.00%) high severe
Benchmarking decode/scalar_2bit/1024
Benchmarking decode/scalar_2bit/1024: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/1024: Collecting 100 samples in estimated 5.0001 s (13M iterations)
Benchmarking decode/scalar_2bit/1024: Analyzing
decode/scalar_2bit/1024 time: [381.41 ns 381.62 ns 381.89 ns]
thrpt: [2.4972 GiB/s 2.4990 GiB/s 2.5004 GiB/s]
change:
time: [−6.1743% −3.7635% −1.9771%] (p = 0.00 < 0.05)
thrpt: [+2.0170% +3.9106% +6.5806%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
2 (2.00%) low mild
7 (7.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_4bit/1024
Benchmarking decode/scalar_4bit/1024: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/1024: Collecting 100 samples in estimated 5.0013 s (14M iterations)
Benchmarking decode/scalar_4bit/1024: Analyzing
decode/scalar_4bit/1024 time: [350.72 ns 351.18 ns 351.76 ns]
thrpt: [2.7112 GiB/s 2.7156 GiB/s 2.7192 GiB/s]
change:
time: [−6.7505% −4.4176% −2.4897%] (p = 0.00 < 0.05)
thrpt: [+2.5532% +4.6218% +7.2392%]
Performance has improved.
Found 17 outliers among 100 measurements (17.00%)
5 (5.00%) high mild
12 (12.00%) high severe
Benchmarking decode/simd_4bit/2048
Benchmarking decode/simd_4bit/2048: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/2048: Collecting 100 samples in estimated 5.0007 s (13M iterations)
Benchmarking decode/simd_4bit/2048: Analyzing
decode/simd_4bit/2048 time: [377.87 ns 379.97 ns 384.12 ns]
thrpt: [4.9655 GiB/s 5.0198 GiB/s 5.0477 GiB/s]
change:
time: [−5.5625% −3.6461% −2.0339%] (p = 0.00 < 0.05)
thrpt: [+2.0761% +3.7841% +5.8901%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
1 (1.00%) low mild
2 (2.00%) high mild
8 (8.00%) high severe
Benchmarking decode/scalar_2bit/2048
Benchmarking decode/scalar_2bit/2048: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/2048: Collecting 100 samples in estimated 5.0033 s (6.7M iterations)
Benchmarking decode/scalar_2bit/2048: Analyzing
decode/scalar_2bit/2048 time: [741.90 ns 747.26 ns 755.39 ns]
thrpt: [2.5250 GiB/s 2.5524 GiB/s 2.5709 GiB/s]
change:
time: [−6.6816% −4.0199% −1.8993%] (p = 0.00 < 0.05)
thrpt: [+1.9361% +4.1883% +7.1600%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
1 (1.00%) low mild
6 (6.00%) high mild
7 (7.00%) high severe
Benchmarking decode/scalar_4bit/2048
Benchmarking decode/scalar_4bit/2048: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/2048: Collecting 100 samples in estimated 5.0031 s (7.3M iterations)
Benchmarking decode/scalar_4bit/2048: Analyzing
decode/scalar_4bit/2048 time: [688.46 ns 688.99 ns 689.70 ns]
thrpt: [2.7655 GiB/s 2.7683 GiB/s 2.7705 GiB/s]
change:
time: [−8.0907% −5.3017% −3.1239%] (p = 0.00 < 0.05)
thrpt: [+3.2246% +5.5985% +8.8029%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
2 (2.00%) low mild
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking decode/simd_4bit/4095
Benchmarking decode/simd_4bit/4095: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/4095: Collecting 100 samples in estimated 5.0021 s (6.8M iterations)
Benchmarking decode/simd_4bit/4095: Analyzing
decode/simd_4bit/4095 time: [726.37 ns 727.13 ns 728.22 ns]
thrpt: [5.2371 GiB/s 5.2449 GiB/s 5.2505 GiB/s]
change:
time: [−6.7046% −3.9436% −2.0955%] (p = 0.00 < 0.05)
thrpt: [+2.1403% +4.1056% +7.1865%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
7 (7.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_2bit/4095
Benchmarking decode/scalar_2bit/4095: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/4095: Collecting 100 samples in estimated 5.0013 s (3.4M iterations)
Benchmarking decode/scalar_2bit/4095: Analyzing
decode/scalar_2bit/4095 time: [1.4725 µs 1.4732 µs 1.4742 µs]
thrpt: [2.5871 GiB/s 2.5887 GiB/s 2.5900 GiB/s]
change:
time: [−8.0642% −5.3659% −3.1148%] (p = 0.00 < 0.05)
thrpt: [+3.2149% +5.6702% +8.7716%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
1 (1.00%) low mild
2 (2.00%) high mild
12 (12.00%) high severe
Benchmarking decode/scalar_4bit/4095
Benchmarking decode/scalar_4bit/4095: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/4095: Collecting 100 samples in estimated 5.0003 s (3.6M iterations)
Benchmarking decode/scalar_4bit/4095: Analyzing
decode/scalar_4bit/4095 time: [1.3718 µs 1.3732 µs 1.3749 µs]
thrpt: [2.7739 GiB/s 2.7773 GiB/s 2.7800 GiB/s]
change:
time: [−8.1882% −5.6707% −3.4842%] (p = 0.00 < 0.05)
thrpt: [+3.6100% +6.0116% +8.9185%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking decode/simd_4bit/4096
Benchmarking decode/simd_4bit/4096: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/4096: Collecting 100 samples in estimated 5.0034 s (6.9M iterations)
Benchmarking decode/simd_4bit/4096: Analyzing
decode/simd_4bit/4096 time: [724.05 ns 724.45 ns 724.93 ns]
thrpt: [5.2622 GiB/s 5.2656 GiB/s 5.2686 GiB/s]
change:
time: [−5.3827% −3.6574% −2.2513%] (p = 0.00 < 0.05)
thrpt: [+2.3032% +3.7962% +5.6889%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
3 (3.00%) high mild
8 (8.00%) high severe
Benchmarking decode/scalar_2bit/4096
Benchmarking decode/scalar_2bit/4096: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/4096: Collecting 100 samples in estimated 5.0008 s (3.4M iterations)
Benchmarking decode/scalar_2bit/4096: Analyzing
decode/scalar_2bit/4096 time: [1.4723 µs 1.4732 µs 1.4743 µs]
thrpt: [2.5875 GiB/s 2.5895 GiB/s 2.5910 GiB/s]
change:
time: [−7.4514% −4.7830% −2.6831%] (p = 0.00 < 0.05)
thrpt: [+2.7571% +5.0233% +8.0513%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking decode/scalar_4bit/4096
Benchmarking decode/scalar_4bit/4096: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/4096: Collecting 100 samples in estimated 5.0066 s (3.7M iterations)
Benchmarking decode/scalar_4bit/4096: Analyzing
decode/scalar_4bit/4096 time: [1.3686 µs 1.3766 µs 1.3887 µs]
thrpt: [2.7469 GiB/s 2.7711 GiB/s 2.7873 GiB/s]
change:
time: [−2.3305% −1.9031% −1.3443%] (p = 0.00 < 0.05)
thrpt: [+1.3626% +1.9400% +2.3861%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking decode/simd_4bit/8192
Benchmarking decode/simd_4bit/8192: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/8192: Collecting 100 samples in estimated 5.0002 s (3.5M iterations)
Benchmarking decode/simd_4bit/8192: Analyzing
decode/simd_4bit/8192 time: [1.4249 µs 1.4261 µs 1.4275 µs]
thrpt: [5.3447 GiB/s 5.3498 GiB/s 5.3542 GiB/s]
change:
time: [−7.5865% −5.2729% −3.3327%] (p = 0.00 < 0.05)
thrpt: [+3.4476% +5.5665% +8.2093%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking decode/scalar_2bit/8192
Benchmarking decode/scalar_2bit/8192: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/8192: Collecting 100 samples in estimated 5.0079 s (1.7M iterations)
Benchmarking decode/scalar_2bit/8192: Analyzing
decode/scalar_2bit/8192 time: [2.9218 µs 2.9233 µs 2.9252 µs]
thrpt: [2.6082 GiB/s 2.6098 GiB/s 2.6112 GiB/s]
change:
time: [−6.2533% −4.1826% −2.6307%] (p = 0.00 < 0.05)
thrpt: [+2.7018% +4.3652% +6.6704%]
Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
2 (2.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_4bit/8192
Benchmarking decode/scalar_4bit/8192: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/8192: Collecting 100 samples in estimated 5.0106 s (1.9M iterations)
Benchmarking decode/scalar_4bit/8192: Analyzing
decode/scalar_4bit/8192 time: [2.6939 µs 2.6970 µs 2.7008 µs]
thrpt: [2.8249 GiB/s 2.8289 GiB/s 2.8321 GiB/s]
change:
time: [−14.033% −7.9536% −3.4872%] (p = 0.00 < 0.05)
thrpt: [+3.6132% +8.6409% +16.324%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
2 (2.00%) low mild
5 (5.00%) high mild
5 (5.00%) high severe
Benchmarking decode/simd_4bit/9999
Benchmarking decode/simd_4bit/9999: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/9999: Collecting 100 samples in estimated 5.0010 s (2.9M iterations)
Benchmarking decode/simd_4bit/9999: Analyzing
decode/simd_4bit/9999 time: [1.7362 µs 1.7380 µs 1.7404 µs]
thrpt: [5.3508 GiB/s 5.3580 GiB/s 5.3637 GiB/s]
change:
time: [−9.4257% −6.2354% −3.6412%] (p = 0.00 < 0.05)
thrpt: [+3.7788% +6.6501% +10.407%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
5 (5.00%) high mild
5 (5.00%) high severe
Benchmarking decode/scalar_2bit/9999
Benchmarking decode/scalar_2bit/9999: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/9999: Collecting 100 samples in estimated 5.0136 s (1.4M iterations)
Benchmarking decode/scalar_2bit/9999: Analyzing
decode/scalar_2bit/9999 time: [3.5626 µs 3.5838 µs 3.6251 µs]
thrpt: [2.5688 GiB/s 2.5984 GiB/s 2.6139 GiB/s]
change:
time: [−8.8694% −5.9123% −3.3702%] (p = 0.00 < 0.05)
thrpt: [+3.4877% +6.2838% +9.7326%]
Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
2 (2.00%) high mild
4 (4.00%) high severe
Benchmarking decode/scalar_4bit/9999
Benchmarking decode/scalar_4bit/9999: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/9999: Collecting 100 samples in estimated 5.0128 s (1.5M iterations)
Benchmarking decode/scalar_4bit/9999: Analyzing
decode/scalar_4bit/9999 time: [3.2829 µs 3.2990 µs 3.3305 µs]
thrpt: [2.7961 GiB/s 2.8228 GiB/s 2.8366 GiB/s]
change:
time: [−7.9427% −5.1150% −2.7547%] (p = 0.00 < 0.05)
thrpt: [+2.8327% +5.3908% +8.6280%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking decode/simd_4bit/10000
Benchmarking decode/simd_4bit/10000: Warming up for 3.0000 s
Benchmarking decode/simd_4bit/10000: Collecting 100 samples in estimated 5.0015 s (2.9M iterations)
Benchmarking decode/simd_4bit/10000: Analyzing
decode/simd_4bit/10000 time: [1.7358 µs 1.7376 µs 1.7398 µs]
thrpt: [5.3531 GiB/s 5.3597 GiB/s 5.3655 GiB/s]
change:
time: [−11.016% −6.8875% −3.5091%] (p = 0.00 < 0.05)
thrpt: [+3.6367% +7.3970% +12.380%]
Performance has improved.
Found 7 outliers among 100 measurements (7.00%)
4 (4.00%) high mild
3 (3.00%) high severe
Benchmarking decode/scalar_2bit/10000
Benchmarking decode/scalar_2bit/10000: Warming up for 3.0000 s
Benchmarking decode/scalar_2bit/10000: Collecting 100 samples in estimated 5.0060 s (1.4M iterations)
Benchmarking decode/scalar_2bit/10000: Analyzing
decode/scalar_2bit/10000
time: [3.5610 µs 3.5642 µs 3.5679 µs]
thrpt: [2.6103 GiB/s 2.6130 GiB/s 2.6153 GiB/s]
change:
time: [−7.9356% −5.2629% −3.1533%] (p = 0.00 < 0.05)
thrpt: [+3.2559% +5.5552% +8.6196%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
7 (7.00%) high mild
3 (3.00%) high severe
Benchmarking decode/scalar_4bit/10000
Benchmarking decode/scalar_4bit/10000: Warming up for 3.0000 s
Benchmarking decode/scalar_4bit/10000: Collecting 100 samples in estimated 5.0039 s (1.5M iterations)
Benchmarking decode/scalar_4bit/10000: Analyzing
decode/scalar_4bit/10000
time: [3.2765 µs 3.2781 µs 3.2800 µs]
thrpt: [2.8394 GiB/s 2.8410 GiB/s 2.8424 GiB/s]
change:
time: [−6.1278% −4.2717% −2.7686%] (p = 0.00 < 0.05)
thrpt: [+2.8474% +4.4623% +6.5278%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) low mild
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/simd_4bit/15
Benchmarking roundtrip/simd_4bit/15: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/15: Collecting 100 samples in estimated 5.0002 s (93M iterations)
Benchmarking roundtrip/simd_4bit/15: Analyzing
roundtrip/simd_4bit/15 time: [53.910 ns 53.980 ns 54.069 ns]
thrpt: [264.57 MiB/s 265.01 MiB/s 265.35 MiB/s]
change:
time: [−7.7050% −5.3254% −3.4065%] (p = 0.00 < 0.05)
thrpt: [+3.5267% +5.6249% +8.3482%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
2 (2.00%) low mild
2 (2.00%) high mild
11 (11.00%) high severe
Benchmarking roundtrip/scalar_2bit/15
Benchmarking roundtrip/scalar_2bit/15: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/15: Collecting 100 samples in estimated 5.0002 s (42M iterations)
Benchmarking roundtrip/scalar_2bit/15: Analyzing
roundtrip/scalar_2bit/15
time: [118.95 ns 119.65 ns 120.72 ns]
thrpt: [118.50 MiB/s 119.56 MiB/s 120.26 MiB/s]
change:
time: [−7.0059% −4.4029% −2.3297%] (p = 0.00 < 0.05)
thrpt: [+2.3853% +4.6057% +7.5337%]
Performance has improved.
Found 16 outliers among 100 measurements (16.00%)
8 (8.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/scalar_4bit/15
Benchmarking roundtrip/scalar_4bit/15: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/15: Collecting 100 samples in estimated 5.0002 s (90M iterations)
Benchmarking roundtrip/scalar_4bit/15: Analyzing
roundtrip/scalar_4bit/15
time: [55.381 ns 55.415 ns 55.458 ns]
thrpt: [257.95 MiB/s 258.14 MiB/s 258.30 MiB/s]
change:
time: [+2.2415% +2.4236% +2.6211%] (p = 0.00 < 0.05)
thrpt: [−2.5541% −2.3662% −2.1923%]
Performance has regressed.
Found 7 outliers among 100 measurements (7.00%)
2 (2.00%) high mild
5 (5.00%) high severe
Benchmarking roundtrip/simd_4bit/16
Benchmarking roundtrip/simd_4bit/16: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/16: Collecting 100 samples in estimated 5.0002 s (96M iterations)
Benchmarking roundtrip/simd_4bit/16: Analyzing
roundtrip/simd_4bit/16 time: [51.837 ns 51.894 ns 51.971 ns]
thrpt: [293.60 MiB/s 294.04 MiB/s 294.36 MiB/s]
change:
time: [+2.7225% +5.4013% +10.523%] (p = 0.01 < 0.05)
thrpt: [−9.5207% −5.1245% −2.6503%]
Performance has regressed.
Found 13 outliers among 100 measurements (13.00%)
2 (2.00%) low mild
6 (6.00%) high mild
5 (5.00%) high severe
Benchmarking roundtrip/scalar_2bit/16
Benchmarking roundtrip/scalar_2bit/16: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/16: Collecting 100 samples in estimated 5.0000 s (66M iterations)
Benchmarking roundtrip/scalar_2bit/16: Analyzing
roundtrip/scalar_2bit/16
time: [75.943 ns 75.993 ns 76.062 ns]
thrpt: [200.61 MiB/s 200.79 MiB/s 200.92 MiB/s]
change:
time: [−2.6522% −2.4586% −2.2770%] (p = 0.00 < 0.05)
thrpt: [+2.3300% +2.5206% +2.7245%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
7 (7.00%) high mild
5 (5.00%) high severe
Benchmarking roundtrip/scalar_4bit/16
Benchmarking roundtrip/scalar_4bit/16: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/16: Collecting 100 samples in estimated 5.0001 s (89M iterations)
Benchmarking roundtrip/scalar_4bit/16: Analyzing
roundtrip/scalar_4bit/16
time: [56.091 ns 56.135 ns 56.192 ns]
thrpt: [271.55 MiB/s 271.82 MiB/s 272.03 MiB/s]
change:
time: [+2.0514% +2.2141% +2.3805%] (p = 0.00 < 0.05)
thrpt: [−2.3251% −2.1661% −2.0102%]
Performance has regressed.
Found 9 outliers among 100 measurements (9.00%)
3 (3.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/simd_4bit/17
Benchmarking roundtrip/simd_4bit/17: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/17: Collecting 100 samples in estimated 5.0001 s (88M iterations)
Benchmarking roundtrip/simd_4bit/17: Analyzing
roundtrip/simd_4bit/17 time: [56.848 ns 56.922 ns 57.021 ns]
thrpt: [284.32 MiB/s 284.82 MiB/s 285.19 MiB/s]
change:
time: [+1.3150% +1.9960% +2.5075%] (p = 0.00 < 0.05)
thrpt: [−2.4462% −1.9569% −1.2979%]
Performance has regressed.
Found 12 outliers among 100 measurements (12.00%)
1 (1.00%) low mild
4 (4.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_2bit/17
Benchmarking roundtrip/scalar_2bit/17: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/17: Collecting 100 samples in estimated 5.0004 s (38M iterations)
Benchmarking roundtrip/scalar_2bit/17: Analyzing
roundtrip/scalar_2bit/17
time: [132.25 ns 133.15 ns 134.57 ns]
thrpt: [120.48 MiB/s 121.76 MiB/s 122.59 MiB/s]
change:
time: [−4.7158% −4.3449% −3.7960%] (p = 0.00 < 0.05)
thrpt: [+3.9458% +4.5423% +4.9492%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
3 (3.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_4bit/17
Benchmarking roundtrip/scalar_4bit/17: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/17: Collecting 100 samples in estimated 5.0000 s (88M iterations)
Benchmarking roundtrip/scalar_4bit/17: Analyzing
roundtrip/scalar_4bit/17
time: [56.716 ns 57.255 ns 58.049 ns]
thrpt: [279.29 MiB/s 283.16 MiB/s 285.85 MiB/s]
change:
time: [−3.3289% −2.7532% −1.9262%] (p = 0.00 < 0.05)
thrpt: [+1.9640% +2.8311% +3.4435%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
2 (2.00%) high mild
13 (13.00%) high severe
Benchmarking roundtrip/simd_4bit/32
Benchmarking roundtrip/simd_4bit/32: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/32: Collecting 100 samples in estimated 5.0002 s (80M iterations)
Benchmarking roundtrip/simd_4bit/32: Analyzing
roundtrip/simd_4bit/32 time: [62.398 ns 62.444 ns 62.494 ns]
thrpt: [488.33 MiB/s 488.72 MiB/s 489.08 MiB/s]
change:
time: [+1.6547% +1.9368% +2.2210%] (p = 0.00 < 0.05)
thrpt: [−2.1728% −1.9000% −1.6278%]
Performance has regressed.
Found 10 outliers among 100 measurements (10.00%)
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/scalar_2bit/32
Benchmarking roundtrip/scalar_2bit/32: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/32: Collecting 100 samples in estimated 5.0000 s (51M iterations)
Benchmarking roundtrip/scalar_2bit/32: Analyzing
roundtrip/scalar_2bit/32
time: [97.372 ns 97.449 ns 97.539 ns]
thrpt: [312.88 MiB/s 313.16 MiB/s 313.41 MiB/s]
change:
time: [−2.3735% −2.1471% −1.9005%] (p = 0.00 < 0.05)
thrpt: [+1.9373% +2.1942% +2.4312%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
3 (3.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/scalar_4bit/32
Benchmarking roundtrip/scalar_4bit/32: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/32: Collecting 100 samples in estimated 5.0000 s (71M iterations)
Benchmarking roundtrip/scalar_4bit/32: Analyzing
roundtrip/scalar_4bit/32
time: [70.706 ns 70.751 ns 70.804 ns]
thrpt: [431.02 MiB/s 431.34 MiB/s 431.61 MiB/s]
change:
time: [−10.142% −7.0660% −4.4600%] (p = 0.00 < 0.05)
thrpt: [+4.6682% +7.6032% +11.287%]
Performance has improved.
Found 7 outliers among 100 measurements (7.00%)
2 (2.00%) high mild
5 (5.00%) high severe
Benchmarking roundtrip/simd_4bit/33
Benchmarking roundtrip/simd_4bit/33: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/33: Collecting 100 samples in estimated 5.0001 s (78M iterations)
Benchmarking roundtrip/simd_4bit/33: Analyzing
roundtrip/simd_4bit/33 time: [63.449 ns 63.536 ns 63.643 ns]
thrpt: [494.50 MiB/s 495.33 MiB/s 496.00 MiB/s]
change:
time: [−18.868% −11.261% −4.7286%] (p = 0.00 < 0.05)
thrpt: [+4.9633% +12.689% +23.256%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
4 (4.00%) high mild
4 (4.00%) high severe
Benchmarking roundtrip/scalar_2bit/33
Benchmarking roundtrip/scalar_2bit/33: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/33: Collecting 100 samples in estimated 5.0002 s (34M iterations)
Benchmarking roundtrip/scalar_2bit/33: Analyzing
roundtrip/scalar_2bit/33
time: [147.53 ns 147.68 ns 147.92 ns]
thrpt: [212.76 MiB/s 213.10 MiB/s 213.31 MiB/s]
change:
time: [−4.7084% −4.5467% −4.3743%] (p = 0.00 < 0.05)
thrpt: [+4.5744% +4.7632% +4.9410%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
2 (2.00%) low mild
6 (6.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_4bit/33
Benchmarking roundtrip/scalar_4bit/33: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/33: Collecting 100 samples in estimated 5.0000 s (63M iterations)
Benchmarking roundtrip/scalar_4bit/33: Analyzing
roundtrip/scalar_4bit/33
time: [78.916 ns 79.512 ns 80.423 ns]
thrpt: [391.32 MiB/s 395.81 MiB/s 398.80 MiB/s]
change:
time: [+1.1626% +2.0470% +2.8257%] (p = 0.00 < 0.05)
thrpt: [−2.7481% −2.0059% −1.1493%]
Performance has regressed.
Found 16 outliers among 100 measurements (16.00%)
8 (8.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/simd_4bit/63
Benchmarking roundtrip/simd_4bit/63: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/63: Collecting 100 samples in estimated 5.0000 s (67M iterations)
Benchmarking roundtrip/simd_4bit/63: Analyzing
roundtrip/simd_4bit/63 time: [74.949 ns 75.016 ns 75.116 ns]
thrpt: [799.85 MiB/s 800.92 MiB/s 801.63 MiB/s]
change:
time: [−7.0672% −6.7600% −6.5358%] (p = 0.00 < 0.05)
thrpt: [+6.9929% +7.2501% +7.6047%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
1 (1.00%) low mild
3 (3.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/scalar_2bit/63
Benchmarking roundtrip/scalar_2bit/63: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/63: Collecting 100 samples in estimated 5.0001 s (29M iterations)
Benchmarking roundtrip/scalar_2bit/63: Analyzing
roundtrip/scalar_2bit/63
time: [170.46 ns 170.94 ns 171.49 ns]
thrpt: [350.36 MiB/s 351.47 MiB/s 352.47 MiB/s]
change:
time: [−4.7045% −4.5072% −4.2961%] (p = 0.00 < 0.05)
thrpt: [+4.4890% +4.7200% +4.9367%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) low mild
3 (3.00%) high mild
9 (9.00%) high severe
Benchmarking roundtrip/scalar_4bit/63
Benchmarking roundtrip/scalar_4bit/63: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/63: Collecting 100 samples in estimated 5.0001 s (47M iterations)
Benchmarking roundtrip/scalar_4bit/63: Analyzing
roundtrip/scalar_4bit/63
time: [105.96 ns 106.07 ns 106.25 ns]
thrpt: [565.49 MiB/s 566.44 MiB/s 567.01 MiB/s]
change:
time: [−8.1951% −4.8405% −2.1673%] (p = 0.00 < 0.05)
thrpt: [+2.2153% +5.0867% +8.9267%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
7 (7.00%) high mild
4 (4.00%) high severe
Benchmarking roundtrip/simd_4bit/64
Benchmarking roundtrip/simd_4bit/64: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/64: Collecting 100 samples in estimated 5.0000 s (71M iterations)
Benchmarking roundtrip/simd_4bit/64: Analyzing
roundtrip/simd_4bit/64 time: [72.096 ns 72.328 ns 72.557 ns]
thrpt: [841.20 MiB/s 843.87 MiB/s 846.58 MiB/s]
change:
time: [−11.904% −9.0189% −6.6104%] (p = 0.00 < 0.05)
thrpt: [+7.0783% +9.9130% +13.512%]
Performance has improved.
Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) high severe
Benchmarking roundtrip/scalar_2bit/64
Benchmarking roundtrip/scalar_2bit/64: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/64: Collecting 100 samples in estimated 5.0001 s (39M iterations)
Benchmarking roundtrip/scalar_2bit/64: Analyzing
roundtrip/scalar_2bit/64
time: [124.61 ns 126.16 ns 128.35 ns]
thrpt: [475.54 MiB/s 483.81 MiB/s 489.79 MiB/s]
change:
time: [−8.6250% −6.4275% −4.5484%] (p = 0.00 < 0.05)
thrpt: [+4.7651% +6.8690% +9.4391%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
2 (2.00%) high mild
11 (11.00%) high severe
Benchmarking roundtrip/scalar_4bit/64
Benchmarking roundtrip/scalar_4bit/64: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/64: Collecting 100 samples in estimated 5.0004 s (47M iterations)
Benchmarking roundtrip/scalar_4bit/64: Analyzing
roundtrip/scalar_4bit/64
time: [106.53 ns 107.62 ns 109.09 ns]
thrpt: [559.48 MiB/s 567.12 MiB/s 572.93 MiB/s]
change:
time: [−6.3169% −3.5101% −1.4877%] (p = 0.00 < 0.05)
thrpt: [+1.5102% +3.6378% +6.7428%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
1 (1.00%) low mild
3 (3.00%) high mild
10 (10.00%) high severe
Benchmarking roundtrip/simd_4bit/127
Benchmarking roundtrip/simd_4bit/127: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/127: Collecting 100 samples in estimated 5.0005 s (48M iterations)
Benchmarking roundtrip/simd_4bit/127: Analyzing
roundtrip/simd_4bit/127 time: [103.42 ns 103.50 ns 103.60 ns]
thrpt: [1.1417 GiB/s 1.1428 GiB/s 1.1437 GiB/s]
change:
time: [−8.8195% −6.8690% −5.2980%] (p = 0.00 < 0.05)
thrpt: [+5.5944% +7.3756% +9.6726%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) low mild
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/scalar_2bit/127
Benchmarking roundtrip/scalar_2bit/127: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/127: Collecting 100 samples in estimated 5.0011 s (21M iterations)
Benchmarking roundtrip/scalar_2bit/127: Analyzing
roundtrip/scalar_2bit/127
time: [235.16 ns 235.32 ns 235.53 ns]
thrpt: [514.22 MiB/s 514.69 MiB/s 515.04 MiB/s]
change:
time: [−7.1063% −5.4602% −4.5169%] (p = 0.00 < 0.05)
thrpt: [+4.7306% +5.7755% +7.6500%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
1 (1.00%) low mild
7 (7.00%) high mild
4 (4.00%) high severe
Benchmarking roundtrip/scalar_4bit/127
Benchmarking roundtrip/scalar_4bit/127: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/127: Collecting 100 samples in estimated 5.0004 s (29M iterations)
Benchmarking roundtrip/scalar_4bit/127: Analyzing
roundtrip/scalar_4bit/127
time: [174.85 ns 174.93 ns 175.02 ns]
thrpt: [692.00 MiB/s 692.38 MiB/s 692.70 MiB/s]
change:
time: [−2.2928% −2.1442% −1.9915%] (p = 0.00 < 0.05)
thrpt: [+2.0320% +2.1912% +2.3466%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking roundtrip/simd_4bit/128
Benchmarking roundtrip/simd_4bit/128: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/128: Collecting 100 samples in estimated 5.0001 s (50M iterations)
Benchmarking roundtrip/simd_4bit/128: Analyzing
roundtrip/simd_4bit/128 time: [99.560 ns 99.679 ns 99.824 ns]
thrpt: [1.1942 GiB/s 1.1959 GiB/s 1.1974 GiB/s]
change:
time: [−4.9233% −4.7340% −4.5410%] (p = 0.00 < 0.05)
thrpt: [+4.7570% +4.9693% +5.1783%]
Performance has improved.
Found 16 outliers among 100 measurements (16.00%)
2 (2.00%) low mild
8 (8.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/scalar_2bit/128
Benchmarking roundtrip/scalar_2bit/128: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/128: Collecting 100 samples in estimated 5.0003 s (26M iterations)
Benchmarking roundtrip/scalar_2bit/128: Analyzing
roundtrip/scalar_2bit/128
time: [191.58 ns 191.74 ns 191.95 ns]
thrpt: [635.94 MiB/s 636.64 MiB/s 637.19 MiB/s]
change:
time: [−5.6373% −4.6106% −3.8728%] (p = 0.00 < 0.05)
thrpt: [+4.0288% +4.8334% +5.9740%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/scalar_4bit/128
Benchmarking roundtrip/scalar_4bit/128: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/128: Collecting 100 samples in estimated 5.0002 s (28M iterations)
Benchmarking roundtrip/scalar_4bit/128: Analyzing
roundtrip/scalar_4bit/128
time: [175.81 ns 176.60 ns 178.06 ns]
thrpt: [685.57 MiB/s 691.24 MiB/s 694.33 MiB/s]
change:
time: [−10.703% −7.1316% −4.0227%] (p = 0.00 < 0.05)
thrpt: [+4.1914% +7.6792% +11.985%]
Performance has improved.
Found 17 outliers among 100 measurements (17.00%)
1 (1.00%) low mild
8 (8.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/simd_4bit/255
Benchmarking roundtrip/simd_4bit/255: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/255: Collecting 100 samples in estimated 5.0004 s (32M iterations)
Benchmarking roundtrip/simd_4bit/255: Analyzing
roundtrip/simd_4bit/255 time: [158.25 ns 158.84 ns 160.01 ns]
thrpt: [1.4842 GiB/s 1.4952 GiB/s 1.5007 GiB/s]
change:
time: [−4.1869% −3.8877% −3.4383%] (p = 0.00 < 0.05)
thrpt: [+3.5607% +4.0449% +4.3698%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking roundtrip/scalar_2bit/255
Benchmarking roundtrip/scalar_2bit/255: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/255: Collecting 100 samples in estimated 5.0010 s (14M iterations)
Benchmarking roundtrip/scalar_2bit/255: Analyzing
roundtrip/scalar_2bit/255
time: [358.83 ns 359.09 ns 359.43 ns]
thrpt: [676.60 MiB/s 677.24 MiB/s 677.73 MiB/s]
change:
time: [−13.764% −9.1705% −5.3748%] (p = 0.00 < 0.05)
thrpt: [+5.6801% +10.096% +15.961%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
1 (1.00%) low mild
5 (5.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/scalar_4bit/255
Benchmarking roundtrip/scalar_4bit/255: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/255: Collecting 100 samples in estimated 5.0006 s (16M iterations)
Benchmarking roundtrip/scalar_4bit/255: Analyzing
roundtrip/scalar_4bit/255
time: [305.20 ns 305.43 ns 305.70 ns]
thrpt: [795.50 MiB/s 796.20 MiB/s 796.80 MiB/s]
change:
time: [−3.3063% −3.1541% −2.9943%] (p = 0.00 < 0.05)
thrpt: [+3.0867% +3.2568% +3.4193%]
Performance has improved.
Found 5 outliers among 100 measurements (5.00%)
3 (3.00%) high mild
2 (2.00%) high severe
Benchmarking roundtrip/simd_4bit/256
Benchmarking roundtrip/simd_4bit/256: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/256: Collecting 100 samples in estimated 5.0000 s (33M iterations)
Benchmarking roundtrip/simd_4bit/256: Analyzing
roundtrip/simd_4bit/256 time: [153.03 ns 153.20 ns 153.42 ns]
thrpt: [1.5540 GiB/s 1.5562 GiB/s 1.5580 GiB/s]
change:
time: [−4.5924% −4.3903% −4.1841%] (p = 0.00 < 0.05)
thrpt: [+4.3668% +4.5919% +4.8134%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
4 (4.00%) high mild
10 (10.00%) high severe
Benchmarking roundtrip/scalar_2bit/256
Benchmarking roundtrip/scalar_2bit/256: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/256: Collecting 100 samples in estimated 5.0003 s (16M iterations)
Benchmarking roundtrip/scalar_2bit/256: Analyzing
roundtrip/scalar_2bit/256
time: [307.69 ns 307.84 ns 308.02 ns]
thrpt: [792.61 MiB/s 793.07 MiB/s 793.46 MiB/s]
change:
time: [−4.8947% −3.9199% −3.3322%] (p = 0.00 < 0.05)
thrpt: [+3.4471% +4.0798% +5.1466%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
2 (2.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_4bit/256
Benchmarking roundtrip/scalar_4bit/256: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/256: Collecting 100 samples in estimated 5.0008 s (16M iterations)
Benchmarking roundtrip/scalar_4bit/256: Analyzing
roundtrip/scalar_4bit/256
time: [303.12 ns 305.62 ns 309.38 ns]
thrpt: [789.12 MiB/s 798.84 MiB/s 805.41 MiB/s]
change:
time: [−3.0633% −2.6241% −1.9973%] (p = 0.00 < 0.05)
thrpt: [+2.0380% +2.6948% +3.1601%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking roundtrip/simd_4bit/512
Benchmarking roundtrip/simd_4bit/512: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/512: Collecting 100 samples in estimated 5.0009 s (19M iterations)
Benchmarking roundtrip/simd_4bit/512: Analyzing
roundtrip/simd_4bit/512 time: [261.13 ns 261.26 ns 261.41 ns]
thrpt: [1.8241 GiB/s 1.8252 GiB/s 1.8260 GiB/s]
change:
time: [−3.3202% −3.1760% −3.0180%] (p = 0.00 < 0.05)
thrpt: [+3.1119% +3.2802% +3.4342%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
2 (2.00%) low mild
1 (1.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_2bit/512
Benchmarking roundtrip/scalar_2bit/512: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/512: Collecting 100 samples in estimated 5.0019 s (8.9M iterations)
Benchmarking roundtrip/scalar_2bit/512: Analyzing
roundtrip/scalar_2bit/512
time: [560.80 ns 561.06 ns 561.36 ns]
thrpt: [869.82 MiB/s 870.29 MiB/s 870.68 MiB/s]
change:
time: [−2.5183% −2.3621% −2.2020%] (p = 0.00 < 0.05)
thrpt: [+2.2516% +2.4192% +2.5833%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
2 (2.00%) low mild
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/scalar_4bit/512
Benchmarking roundtrip/scalar_4bit/512: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/512: Collecting 100 samples in estimated 5.0011 s (9.2M iterations)
Benchmarking roundtrip/scalar_4bit/512: Analyzing
roundtrip/scalar_4bit/512
time: [541.77 ns 542.06 ns 542.40 ns]
thrpt: [900.22 MiB/s 900.79 MiB/s 901.27 MiB/s]
change:
time: [−2.5389% −2.3674% −2.2023%] (p = 0.00 < 0.05)
thrpt: [+2.2519% +2.4248% +2.6050%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
5 (5.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/simd_4bit/1023
Benchmarking roundtrip/simd_4bit/1023: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/1023: Collecting 100 samples in estimated 5.0009 s (10M iterations)
Benchmarking roundtrip/simd_4bit/1023: Analyzing
roundtrip/simd_4bit/1023
time: [478.39 ns 478.95 ns 479.72 ns]
thrpt: [1.9860 GiB/s 1.9892 GiB/s 1.9915 GiB/s]
change:
time: [−3.2619% −3.0424% −2.8188%] (p = 0.00 < 0.05)
thrpt: [+2.9006% +3.1379% +3.3719%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/scalar_2bit/1023
Benchmarking roundtrip/scalar_2bit/1023: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/1023: Collecting 100 samples in estimated 5.0027 s (4.5M iterations)
Benchmarking roundtrip/scalar_2bit/1023: Analyzing
roundtrip/scalar_2bit/1023
time: [1.1175 µs 1.1181 µs 1.1188 µs]
thrpt: [872.03 MiB/s 872.58 MiB/s 873.06 MiB/s]
change:
time: [−3.5372% −3.1313% −2.8423%] (p = 0.00 < 0.05)
thrpt: [+2.9254% +3.2325% +3.6669%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
3 (3.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_4bit/1023
Benchmarking roundtrip/scalar_4bit/1023: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/1023: Collecting 100 samples in estimated 5.0028 s (4.8M iterations)
Benchmarking roundtrip/scalar_4bit/1023: Analyzing
roundtrip/scalar_4bit/1023
time: [1.0496 µs 1.0592 µs 1.0735 µs]
thrpt: [908.82 MiB/s 921.07 MiB/s 929.50 MiB/s]
change:
time: [−6.2725% −3.6062% −1.6738%] (p = 0.00 < 0.05)
thrpt: [+1.7023% +3.7411% +6.6923%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
5 (5.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/simd_4bit/1024
Benchmarking roundtrip/simd_4bit/1024: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/1024: Collecting 100 samples in estimated 5.0022 s (11M iterations)
Benchmarking roundtrip/simd_4bit/1024: Analyzing
roundtrip/simd_4bit/1024
time: [474.43 ns 474.98 ns 475.68 ns]
thrpt: [2.0049 GiB/s 2.0078 GiB/s 2.0102 GiB/s]
change:
time: [−6.5018% −4.5857% −3.0778%] (p = 0.00 < 0.05)
thrpt: [+3.1756% +4.8061% +6.9539%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
2 (2.00%) high mild
9 (9.00%) high severe
Benchmarking roundtrip/scalar_2bit/1024
Benchmarking roundtrip/scalar_2bit/1024: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/1024: Collecting 100 samples in estimated 5.0046 s (4.8M iterations)
Benchmarking roundtrip/scalar_2bit/1024: Analyzing
roundtrip/scalar_2bit/1024
time: [1.0308 µs 1.0314 µs 1.0322 µs]
thrpt: [946.06 MiB/s 946.82 MiB/s 947.40 MiB/s]
change:
time: [−6.7099% −4.5757% −2.8540%] (p = 0.00 < 0.05)
thrpt: [+2.9379% +4.7951% +7.1925%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
8 (8.00%) high mild
4 (4.00%) high severe
Benchmarking roundtrip/scalar_4bit/1024
Benchmarking roundtrip/scalar_4bit/1024: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/1024: Collecting 100 samples in estimated 5.0028 s (4.8M iterations)
Benchmarking roundtrip/scalar_4bit/1024: Analyzing
roundtrip/scalar_4bit/1024
time: [1.0461 µs 1.0466 µs 1.0473 µs]
thrpt: [932.45 MiB/s 933.08 MiB/s 933.57 MiB/s]
change:
time: [−6.6791% −4.1979% −2.4498%] (p = 0.00 < 0.05)
thrpt: [+2.5113% +4.3819% +7.1571%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/simd_4bit/2048
Benchmarking roundtrip/simd_4bit/2048: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/2048: Collecting 100 samples in estimated 5.0018 s (5.4M iterations)
Benchmarking roundtrip/simd_4bit/2048: Analyzing
roundtrip/simd_4bit/2048
time: [923.47 ns 926.56 ns 932.53 ns]
thrpt: [2.0454 GiB/s 2.0585 GiB/s 2.0654 GiB/s]
change:
time: [−6.5838% −3.7132% −1.2014%] (p = 0.00 < 0.05)
thrpt: [+1.2160% +3.8564% +7.0478%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
6 (6.00%) high mild
6 (6.00%) high severe
Benchmarking roundtrip/scalar_2bit/2048
Benchmarking roundtrip/scalar_2bit/2048: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/2048: Collecting 100 samples in estimated 5.0054 s (2.5M iterations)
Benchmarking roundtrip/scalar_2bit/2048: Analyzing
roundtrip/scalar_2bit/2048
time: [1.9897 µs 1.9915 µs 1.9938 µs]
thrpt: [979.59 MiB/s 980.75 MiB/s 981.62 MiB/s]
change:
time: [−10.647% −6.6346% −3.3597%] (p = 0.00 < 0.05)
thrpt: [+3.4764% +7.1061% +11.916%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
2 (2.00%) low mild
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/scalar_4bit/2048
Benchmarking roundtrip/scalar_4bit/2048: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/2048: Collecting 100 samples in estimated 5.0012 s (2.5M iterations)
Benchmarking roundtrip/scalar_4bit/2048: Analyzing
roundtrip/scalar_4bit/2048
time: [2.0203 µs 2.0218 µs 2.0240 µs]
thrpt: [964.96 MiB/s 966.03 MiB/s 966.77 MiB/s]
change:
time: [−2.7692% −2.5898% −2.4168%] (p = 0.00 < 0.05)
thrpt: [+2.4766% +2.6587% +2.8481%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
7 (7.00%) high mild
5 (5.00%) high severe
Benchmarking roundtrip/simd_4bit/4095
Benchmarking roundtrip/simd_4bit/4095: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/4095: Collecting 100 samples in estimated 5.0009 s (2.7M iterations)
Benchmarking roundtrip/simd_4bit/4095: Analyzing
roundtrip/simd_4bit/4095
time: [1.8100 µs 1.8119 µs 1.8144 µs]
thrpt: [2.1020 GiB/s 2.1049 GiB/s 2.1070 GiB/s]
change:
time: [−2.5755% −2.4443% −2.2969%] (p = 0.00 < 0.05)
thrpt: [+2.3509% +2.5055% +2.6436%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
6 (6.00%) high mild
2 (2.00%) high severe
Benchmarking roundtrip/scalar_2bit/4095
Benchmarking roundtrip/scalar_2bit/4095: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/4095: Collecting 100 samples in estimated 5.0130 s (1.2M iterations)
Benchmarking roundtrip/scalar_2bit/4095: Analyzing
roundtrip/scalar_2bit/4095
time: [4.0162 µs 4.0291 µs 4.0530 µs]
thrpt: [963.55 MiB/s 969.26 MiB/s 972.39 MiB/s]
change:
time: [−2.2906% −2.0359% −1.6631%] (p = 0.00 < 0.05)
thrpt: [+1.6913% +2.0782% +2.3443%]
Performance has improved.
Found 14 outliers among 100 measurements (14.00%)
2 (2.00%) low mild
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/scalar_4bit/4095
Benchmarking roundtrip/scalar_4bit/4095: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/4095: Collecting 100 samples in estimated 5.0017 s (1.3M iterations)
Benchmarking roundtrip/scalar_4bit/4095: Analyzing
roundtrip/scalar_4bit/4095
time: [3.9885 µs 3.9918 µs 3.9962 µs]
thrpt: [977.24 MiB/s 978.33 MiB/s 979.14 MiB/s]
change:
time: [−2.5704% −2.3896% −2.2164%] (p = 0.00 < 0.05)
thrpt: [+2.2666% +2.4481% +2.6382%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking roundtrip/simd_4bit/4096
Benchmarking roundtrip/simd_4bit/4096: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/4096: Collecting 100 samples in estimated 5.0034 s (2.8M iterations)
Benchmarking roundtrip/simd_4bit/4096: Analyzing
roundtrip/simd_4bit/4096
time: [1.8070 µs 1.8083 µs 1.8100 µs]
thrpt: [2.1076 GiB/s 2.1095 GiB/s 2.1111 GiB/s]
change:
time: [−2.9676% −2.7009% −2.4610%] (p = 0.00 < 0.05)
thrpt: [+2.5231% +2.7759% +3.0584%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
5 (5.00%) high mild
3 (3.00%) high severe
Benchmarking roundtrip/scalar_2bit/4096
Benchmarking roundtrip/scalar_2bit/4096: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/4096: Collecting 100 samples in estimated 5.0129 s (1.3M iterations)
Benchmarking roundtrip/scalar_2bit/4096: Analyzing
roundtrip/scalar_2bit/4096
time: [3.9031 µs 3.9070 µs 3.9118 µs]
thrpt: [998.59 MiB/s 999.82 MiB/s 1000.8 MiB/s]
change:
time: [−3.8001% −2.8221% −2.2519%] (p = 0.00 < 0.05)
thrpt: [+2.3038% +2.9041% +3.9502%]
Performance has improved.
Found 12 outliers among 100 measurements (12.00%)
1 (1.00%) low mild
4 (4.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_4bit/4096
Benchmarking roundtrip/scalar_4bit/4096: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/4096: Collecting 100 samples in estimated 5.0181 s (1.3M iterations)
Benchmarking roundtrip/scalar_4bit/4096: Analyzing
roundtrip/scalar_4bit/4096
time: [3.9869 µs 3.9988 µs 4.0175 µs]
thrpt: [972.31 MiB/s 976.84 MiB/s 979.78 MiB/s]
change:
time: [−21.373% −13.781% −6.7724%] (p = 0.00 < 0.05)
thrpt: [+7.2644% +15.983% +27.183%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
3 (3.00%) high mild
8 (8.00%) high severe
Benchmarking roundtrip/simd_4bit/8192
Benchmarking roundtrip/simd_4bit/8192: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/8192: Collecting 100 samples in estimated 5.0168 s (1.4M iterations)
Benchmarking roundtrip/simd_4bit/8192: Analyzing
roundtrip/simd_4bit/8192
time: [3.5668 µs 3.5721 µs 3.5789 µs]
thrpt: [2.1318 GiB/s 2.1359 GiB/s 2.1390 GiB/s]
change:
time: [−2.6543% −2.4122% −2.1469%] (p = 0.00 < 0.05)
thrpt: [+2.1940% +2.4718% +2.7267%]
Performance has improved.
Found 17 outliers among 100 measurements (17.00%)
8 (8.00%) high mild
9 (9.00%) high severe
Benchmarking roundtrip/scalar_2bit/8192
Benchmarking roundtrip/scalar_2bit/8192: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/8192: Collecting 100 samples in estimated 5.0025 s (651k iterations)
Benchmarking roundtrip/scalar_2bit/8192: Analyzing
roundtrip/scalar_2bit/8192
time: [7.6700 µs 7.6753 µs 7.6822 µs]
thrpt: [1017.0 MiB/s 1017.9 MiB/s 1018.6 MiB/s]
change:
time: [−8.8286% −5.7800% −3.2631%] (p = 0.00 < 0.05)
thrpt: [+3.3731% +6.1346% +9.6835%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
3 (3.00%) low mild
5 (5.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_4bit/8192
Benchmarking roundtrip/scalar_4bit/8192: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/8192: Collecting 100 samples in estimated 5.0207 s (636k iterations)
Benchmarking roundtrip/scalar_4bit/8192: Analyzing
roundtrip/scalar_4bit/8192
time: [7.8804 µs 7.8858 µs 7.8937 µs]
thrpt: [989.71 MiB/s 990.71 MiB/s 991.38 MiB/s]
change:
time: [−2.6371% −2.4659% −2.2830%] (p = 0.00 < 0.05)
thrpt: [+2.3363% +2.5282% +2.7085%]
Performance has improved.
Found 7 outliers among 100 measurements (7.00%)
5 (5.00%) high mild
2 (2.00%) high severe
Benchmarking roundtrip/simd_4bit/9999
Benchmarking roundtrip/simd_4bit/9999: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/9999: Collecting 100 samples in estimated 5.0214 s (1.2M iterations)
Benchmarking roundtrip/simd_4bit/9999: Analyzing
roundtrip/simd_4bit/9999
time: [4.3553 µs 4.3590 µs 4.3638 µs]
thrpt: [2.1340 GiB/s 2.1363 GiB/s 2.1381 GiB/s]
change:
time: [−2.9188% −2.7449% −2.5597%] (p = 0.00 < 0.05)
thrpt: [+2.6269% +2.8224% +3.0065%]
Performance has improved.
Found 7 outliers among 100 measurements (7.00%)
5 (5.00%) high mild
2 (2.00%) high severe
Benchmarking roundtrip/scalar_2bit/9999
Benchmarking roundtrip/scalar_2bit/9999: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/9999: Collecting 100 samples in estimated 5.0072 s (525k iterations)
Benchmarking roundtrip/scalar_2bit/9999: Analyzing
roundtrip/scalar_2bit/9999
time: [9.5244 µs 9.5360 µs 9.5508 µs]
thrpt: [998.43 MiB/s 999.98 MiB/s 1001.2 MiB/s]
change:
time: [−3.6346% −2.9437% −2.5018%] (p = 0.00 < 0.05)
thrpt: [+2.5660% +3.0330% +3.7717%]
Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
2 (2.00%) low mild
4 (4.00%) high mild
9 (9.00%) high severe
Benchmarking roundtrip/scalar_4bit/9999
Benchmarking roundtrip/scalar_4bit/9999: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/9999: Collecting 100 samples in estimated 5.0075 s (520k iterations)
Benchmarking roundtrip/scalar_4bit/9999: Analyzing
roundtrip/scalar_4bit/9999
time: [9.6152 µs 9.6416 µs 9.6885 µs]
thrpt: [984.24 MiB/s 989.03 MiB/s 991.74 MiB/s]
change:
time: [−2.5921% −2.0435% −1.1772%] (p = 0.00 < 0.05)
thrpt: [+1.1913% +2.0861% +2.6611%]
Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
8 (8.00%) high severe
Benchmarking roundtrip/simd_4bit/10000
Benchmarking roundtrip/simd_4bit/10000: Warming up for 3.0000 s
Benchmarking roundtrip/simd_4bit/10000: Collecting 100 samples in estimated 5.0188 s (1.1M iterations)
Benchmarking roundtrip/simd_4bit/10000: Analyzing
roundtrip/simd_4bit/10000
time: [4.3508 µs 4.3540 µs 4.3579 µs]
thrpt: [2.1371 GiB/s 2.1390 GiB/s 2.1406 GiB/s]
change:
time: [−3.0323% −2.8465% −2.6468%] (p = 0.00 < 0.05)
thrpt: [+2.7188% +2.9299% +3.1271%]
Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
5 (5.00%) high mild
5 (5.00%) high severe
Benchmarking roundtrip/scalar_2bit/10000
Benchmarking roundtrip/scalar_2bit/10000: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_2bit/10000: Collecting 100 samples in estimated 5.0024 s (535k iterations)
Benchmarking roundtrip/scalar_2bit/10000: Analyzing
roundtrip/scalar_2bit/10000
time: [9.3328 µs 9.3389 µs 9.3464 µs]
thrpt: [1020.4 MiB/s 1021.2 MiB/s 1021.9 MiB/s]
change:
time: [−2.8071% −2.6646% −2.5203%] (p = 0.00 < 0.05)
thrpt: [+2.5855% +2.7375% +2.8882%]
Performance has improved.
Found 13 outliers among 100 measurements (13.00%)
6 (6.00%) high mild
7 (7.00%) high severe
Benchmarking roundtrip/scalar_4bit/10000
Benchmarking roundtrip/scalar_4bit/10000: Warming up for 3.0000 s
Benchmarking roundtrip/scalar_4bit/10000: Collecting 100 samples in estimated 5.0069 s (520k iterations)
Benchmarking roundtrip/scalar_4bit/10000: Analyzing
roundtrip/scalar_4bit/10000
time: [9.6174 µs 9.6231 µs 9.6294 µs]
thrpt: [990.38 MiB/s 991.02 MiB/s 991.61 MiB/s]
change:
time: [−2.7167% −2.5029% −2.3205%] (p = 0.00 < 0.05)
thrpt: [+2.3756% +2.5672% +2.7926%]
Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/simd_high_level/15
Benchmarking reverse_complement/simd_high_level/15: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/15: Collecting 100 samples in estimated 5.0001 s (56M iterations)
Benchmarking reverse_complement/simd_high_level/15: Analyzing
reverse_complement/simd_high_level/15
time: [89.812 ns 89.884 ns 89.960 ns]
thrpt: [159.02 MiB/s 159.15 MiB/s 159.28 MiB/s]
change:
time: [−2.6407% −2.4855% −2.3224%] (p = 0.00 < 0.05)
thrpt: [+2.3776% +2.5489% +2.7123%]
Performance has improved.
Found 5 outliers among 100 measurements (5.00%)
1 (1.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/simd_encoded/15
Benchmarking reverse_complement/simd_encoded/15: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/15: Collecting 100 samples in estimated 5.0000 s (127M iterations)
Benchmarking reverse_complement/simd_encoded/15: Analyzing
reverse_complement/simd_encoded/15
time: [39.278 ns 39.352 ns 39.429 ns]
thrpt: [362.81 MiB/s 363.52 MiB/s 364.20 MiB/s]
change:
time: [−0.8605% −0.2349% +0.2516%] (p = 0.47 > 0.05)
thrpt: [−0.2509% +0.2355% +0.8680%]
No change in performance detected.
Found 5 outliers among 100 measurements (5.00%)
1 (1.00%) low mild
4 (4.00%) high mild
Benchmarking reverse_complement/scalar/15
Benchmarking reverse_complement/scalar/15: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/15: Collecting 100 samples in estimated 5.0001 s (202M iterations)
Benchmarking reverse_complement/scalar/15: Analyzing
reverse_complement/scalar/15
time: [24.739 ns 24.826 ns 24.975 ns]
thrpt: [572.77 MiB/s 576.22 MiB/s 578.25 MiB/s]
change:
time: [−13.339% −12.610% −11.868%] (p = 0.00 < 0.05)
thrpt: [+13.467% +14.430% +15.392%]
Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
4 (4.00%) high mild
7 (7.00%) high severe
Benchmarking reverse_complement/simd_high_level/16
Benchmarking reverse_complement/simd_high_level/16: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/16: Collecting 100 samples in estimated 5.0002 s (71M iterations)
Benchmarking reverse_complement/simd_high_level/16: Analyzing
reverse_complement/simd_high_level/16
time: [70.135 ns 70.186 ns 70.246 ns]
thrpt: [217.22 MiB/s 217.41 MiB/s 217.56 MiB/s]
change:
time: [+0.0202% +5.9819% +13.878%] (p = 0.09 > 0.05)
thrpt: [−12.187% −5.6442% −0.0202%]
No change in performance detected.
Found 12 outliers among 100 measurements (12.00%)
2 (2.00%) high mild
10 (10.00%) high severe
Benchmarking reverse_complement/simd_encoded/16
Benchmarking reverse_complement/simd_encoded/16: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/16: Collecting 100 samples in estimated 5.0001 s (224M iterations)
Benchmarking reverse_complement/simd_encoded/16: Analyzing
reverse_complement/simd_encoded/16
time: [22.311 ns 22.329 ns 22.355 ns]
thrpt: [682.58 MiB/s 683.37 MiB/s 683.91 MiB/s]
change:
time: [−0.9252% −0.7712% −0.6172%] (p = 0.00 < 0.05)
thrpt: [+0.6210% +0.7772% +0.9338%]
Change within noise threshold.
Found 9 outliers among 100 measurements (9.00%)
1 (1.00%) low mild
4 (4.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/scalar/16
Benchmarking reverse_complement/scalar/16: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/16: Collecting 100 samples in estimated 5.0001 s (207M iterations)
Benchmarking reverse_complement/scalar/16: Analyzing
reverse_complement/scalar/16
time: [24.181 ns 24.189 ns 24.197 ns]
thrpt: [630.59 MiB/s 630.82 MiB/s 631.02 MiB/s]
change:
time: [+0.0675% +0.2949% +0.5355%] (p = 0.01 < 0.05)
thrpt: [−0.5327% −0.2940% −0.0675%]
Change within noise threshold.
Found 14 outliers among 100 measurements (14.00%)
3 (3.00%) high mild
11 (11.00%) high severe
Benchmarking reverse_complement/simd_high_level/17
Benchmarking reverse_complement/simd_high_level/17: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/17: Collecting 100 samples in estimated 5.0002 s (53M iterations)
Benchmarking reverse_complement/simd_high_level/17: Analyzing
reverse_complement/simd_high_level/17
time: [94.386 ns 94.457 ns 94.551 ns]
thrpt: [171.47 MiB/s 171.64 MiB/s 171.77 MiB/s]
change:
time: [−0.8847% −0.6584% −0.4341%] (p = 0.00 < 0.05)
thrpt: [+0.4360% +0.6627% +0.8926%]
Change within noise threshold.
Found 7 outliers among 100 measurements (7.00%)
3 (3.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/simd_encoded/17
Benchmarking reverse_complement/simd_encoded/17: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/17: Collecting 100 samples in estimated 5.0002 s (119M iterations)
Benchmarking reverse_complement/simd_encoded/17: Analyzing
reverse_complement/simd_encoded/17
time: [42.037 ns 42.070 ns 42.124 ns]
thrpt: [384.88 MiB/s 385.37 MiB/s 385.67 MiB/s]
change:
time: [−0.3475% −0.1359% +0.0671%] (p = 0.21 > 0.05)
thrpt: [−0.0670% +0.1361% +0.3487%]
No change in performance detected.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/scalar/17
Benchmarking reverse_complement/scalar/17: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/17: Collecting 100 samples in estimated 5.0001 s (184M iterations)
Benchmarking reverse_complement/scalar/17: Analyzing
reverse_complement/scalar/17
time: [27.051 ns 27.079 ns 27.114 ns]
thrpt: [597.95 MiB/s 598.71 MiB/s 599.33 MiB/s]
change:
time: [−0.0443% +0.1362% +0.3068%] (p = 0.13 > 0.05)
thrpt: [−0.3059% −0.1360% +0.0443%]
No change in performance detected.
Found 12 outliers among 100 measurements (12.00%)
4 (4.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/simd_high_level/32
Benchmarking reverse_complement/simd_high_level/32: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/32: Collecting 100 samples in estimated 5.0002 s (64M iterations)
Benchmarking reverse_complement/simd_high_level/32: Analyzing
reverse_complement/simd_high_level/32
time: [78.237 ns 78.885 ns 79.854 ns]
thrpt: [382.17 MiB/s 386.86 MiB/s 390.07 MiB/s]
change:
time: [+0.0317% +0.6020% +1.3603%] (p = 0.07 > 0.05)
thrpt: [−1.3421% −0.5984% −0.0317%]
No change in performance detected.
Found 12 outliers among 100 measurements (12.00%)
6 (6.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/simd_encoded/32
Benchmarking reverse_complement/simd_encoded/32: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/32: Collecting 100 samples in estimated 5.0001 s (221M iterations)
Benchmarking reverse_complement/simd_encoded/32: Analyzing
reverse_complement/simd_encoded/32
time: [22.591 ns 22.606 ns 22.625 ns]
thrpt: [1.3172 GiB/s 1.3183 GiB/s 1.3192 GiB/s]
change:
time: [−0.7645% −0.6327% −0.5081%] (p = 0.00 < 0.05)
thrpt: [+0.5107% +0.6367% +0.7704%]
Change within noise threshold.
Found 6 outliers among 100 measurements (6.00%)
2 (2.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/scalar/32
Benchmarking reverse_complement/scalar/32: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/32: Collecting 100 samples in estimated 5.0000 s (149M iterations)
Benchmarking reverse_complement/scalar/32: Analyzing
reverse_complement/scalar/32
time: [33.531 ns 33.552 ns 33.578 ns]
thrpt: [908.87 MiB/s 909.57 MiB/s 910.12 MiB/s]
change:
time: [−0.2281% −0.0525% +0.1055%] (p = 0.55 > 0.05)
thrpt: [−0.1053% +0.0525% +0.2286%]
No change in performance detected.
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/33
Benchmarking reverse_complement/simd_high_level/33: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/33: Collecting 100 samples in estimated 5.0006 s (40M iterations)
Benchmarking reverse_complement/simd_high_level/33: Analyzing
reverse_complement/simd_high_level/33
time: [125.96 ns 126.06 ns 126.18 ns]
thrpt: [249.42 MiB/s 249.65 MiB/s 249.85 MiB/s]
change:
time: [−0.2320% +0.0157% +0.2575%] (p = 0.90 > 0.05)
thrpt: [−0.2569% −0.0157% +0.2325%]
No change in performance detected.
Found 9 outliers among 100 measurements (9.00%)
1 (1.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/simd_encoded/33
Benchmarking reverse_complement/simd_encoded/33: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/33: Collecting 100 samples in estimated 5.0001 s (76M iterations)
Benchmarking reverse_complement/simd_encoded/33: Analyzing
reverse_complement/simd_encoded/33
time: [65.486 ns 65.680 ns 65.960 ns]
thrpt: [477.13 MiB/s 479.16 MiB/s 480.58 MiB/s]
change:
time: [+0.0793% +0.6627% +1.7372%] (p = 0.09 > 0.05)
thrpt: [−1.7075% −0.6583% −0.0792%]
No change in performance detected.
Found 10 outliers among 100 measurements (10.00%)
2 (2.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/scalar/33
Benchmarking reverse_complement/scalar/33: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/33: Collecting 100 samples in estimated 5.0000 s (145M iterations)
Benchmarking reverse_complement/scalar/33: Analyzing
reverse_complement/scalar/33
time: [34.391 ns 34.415 ns 34.449 ns]
thrpt: [913.56 MiB/s 914.45 MiB/s 915.11 MiB/s]
change:
time: [−0.3077% −0.1656% −0.0216%] (p = 0.02 < 0.05)
thrpt: [+0.0216% +0.1659% +0.3086%]
Change within noise threshold.
Found 7 outliers among 100 measurements (7.00%)
2 (2.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/63
Benchmarking reverse_complement/simd_high_level/63: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/63: Collecting 100 samples in estimated 5.0004 s (43M iterations)
Benchmarking reverse_complement/simd_high_level/63: Analyzing
reverse_complement/simd_high_level/63
time: [115.94 ns 116.50 ns 117.55 ns]
thrpt: [511.12 MiB/s 515.71 MiB/s 518.19 MiB/s]
change:
time: [−0.2204% +0.0510% +0.4619%] (p = 0.78 > 0.05)
thrpt: [−0.4598% −0.0510% +0.2209%]
No change in performance detected.
Found 9 outliers among 100 measurements (9.00%)
1 (1.00%) low mild
4 (4.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/simd_encoded/63
Benchmarking reverse_complement/simd_encoded/63: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/63: Collecting 100 samples in estimated 5.0002 s (122M iterations)
Benchmarking reverse_complement/simd_encoded/63: Analyzing
reverse_complement/simd_encoded/63
time: [40.992 ns 41.092 ns 41.229 ns]
thrpt: [1.4231 GiB/s 1.4279 GiB/s 1.4313 GiB/s]
change:
time: [−0.0815% +0.2028% +0.4840%] (p = 0.16 > 0.05)
thrpt: [−0.4817% −0.2024% +0.0816%]
No change in performance detected.
Found 5 outliers among 100 measurements (5.00%)
3 (3.00%) high mild
2 (2.00%) high severe
Benchmarking reverse_complement/scalar/63
Benchmarking reverse_complement/scalar/63: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/63: Collecting 100 samples in estimated 5.0002 s (111M iterations)
Benchmarking reverse_complement/scalar/63: Analyzing
reverse_complement/scalar/63
time: [45.153 ns 45.194 ns 45.246 ns]
thrpt: [1.2968 GiB/s 1.2983 GiB/s 1.2994 GiB/s]
change:
time: [−0.1244% +0.0456% +0.2158%] (p = 0.60 > 0.05)
thrpt: [−0.2153% −0.0456% +0.1245%]
No change in performance detected.
Found 13 outliers among 100 measurements (13.00%)
8 (8.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/64
Benchmarking reverse_complement/simd_high_level/64: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/64: Collecting 100 samples in estimated 5.0002 s (52M iterations)
Benchmarking reverse_complement/simd_high_level/64: Analyzing
reverse_complement/simd_high_level/64
time: [96.174 ns 96.269 ns 96.378 ns]
thrpt: [633.29 MiB/s 634.01 MiB/s 634.64 MiB/s]
change:
time: [−0.1711% +0.0182% +0.2174%] (p = 0.85 > 0.05)
thrpt: [−0.2169% −0.0182% +0.1714%]
No change in performance detected.
Found 5 outliers among 100 measurements (5.00%)
3 (3.00%) high mild
2 (2.00%) high severe
Benchmarking reverse_complement/simd_encoded/64
Benchmarking reverse_complement/simd_encoded/64: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/64: Collecting 100 samples in estimated 5.0001 s (198M iterations)
Benchmarking reverse_complement/simd_encoded/64: Analyzing
reverse_complement/simd_encoded/64
time: [25.161 ns 25.209 ns 25.267 ns]
thrpt: [2.3590 GiB/s 2.3644 GiB/s 2.3689 GiB/s]
change:
time: [−0.0646% +0.4229% +0.9251%] (p = 0.09 > 0.05)
thrpt: [−0.9166% −0.4211% +0.0647%]
No change in performance detected.
Found 15 outliers among 100 measurements (15.00%)
10 (10.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/scalar/64
Benchmarking reverse_complement/scalar/64: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/64: Collecting 100 samples in estimated 5.0002 s (110M iterations)
Benchmarking reverse_complement/scalar/64: Analyzing
reverse_complement/scalar/64
time: [45.308 ns 45.360 ns 45.424 ns]
thrpt: [1.3122 GiB/s 1.3140 GiB/s 1.3156 GiB/s]
change:
time: [+0.3972% +1.3174% +1.9997%] (p = 0.00 < 0.05)
thrpt: [−1.9605% −1.3002% −0.3956%]
Change within noise threshold.
Found 14 outliers among 100 measurements (14.00%)
5 (5.00%) high mild
9 (9.00%) high severe
Benchmarking reverse_complement/simd_high_level/127
Benchmarking reverse_complement/simd_high_level/127: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/127: Collecting 100 samples in estimated 5.0007 s (33M iterations)
Benchmarking reverse_complement/simd_high_level/127: Analyzing
reverse_complement/simd_high_level/127
time: [152.91 ns 153.49 ns 154.33 ns]
thrpt: [784.81 MiB/s 789.08 MiB/s 792.09 MiB/s]
change:
time: [−2.3606% −0.5091% +1.1146%] (p = 0.60 > 0.05)
thrpt: [−1.1023% +0.5117% +2.4177%]
No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
1 (1.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/simd_encoded/127
Benchmarking reverse_complement/simd_encoded/127: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/127: Collecting 100 samples in estimated 5.0000 s (93M iterations)
Benchmarking reverse_complement/simd_encoded/127: Analyzing
reverse_complement/simd_encoded/127
time: [53.556 ns 53.693 ns 53.883 ns]
thrpt: [2.1951 GiB/s 2.2029 GiB/s 2.2085 GiB/s]
change:
time: [−0.3564% +0.4306% +1.7590%] (p = 0.51 > 0.05)
thrpt: [−1.7286% −0.4288% +0.3576%]
No change in performance detected.
Found 13 outliers among 100 measurements (13.00%)
6 (6.00%) high mild
7 (7.00%) high severe
Benchmarking reverse_complement/scalar/127
Benchmarking reverse_complement/scalar/127: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/127: Collecting 100 samples in estimated 5.0001 s (78M iterations)
Benchmarking reverse_complement/scalar/127: Analyzing
reverse_complement/scalar/127
time: [64.142 ns 64.269 ns 64.434 ns]
thrpt: [1.8356 GiB/s 1.8404 GiB/s 1.8440 GiB/s]
change:
time: [−0.0062% +0.1740% +0.3603%] (p = 0.07 > 0.05)
thrpt: [−0.3590% −0.1737% +0.0062%]
No change in performance detected.
Found 8 outliers among 100 measurements (8.00%)
4 (4.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/simd_high_level/128
Benchmarking reverse_complement/simd_high_level/128: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/128: Collecting 100 samples in estimated 5.0002 s (41M iterations)
Benchmarking reverse_complement/simd_high_level/128: Analyzing
reverse_complement/simd_high_level/128
time: [122.37 ns 122.49 ns 122.65 ns]
thrpt: [995.30 MiB/s 996.55 MiB/s 997.54 MiB/s]
change:
time: [−0.4436% −0.2384% −0.0443%] (p = 0.02 < 0.05)
thrpt: [+0.0444% +0.2390% +0.4456%]
Change within noise threshold.
Found 8 outliers among 100 measurements (8.00%)
5 (5.00%) high mild
3 (3.00%) high severe
Benchmarking reverse_complement/simd_encoded/128
Benchmarking reverse_complement/simd_encoded/128: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/128: Collecting 100 samples in estimated 5.0000 s (199M iterations)
Benchmarking reverse_complement/simd_encoded/128: Analyzing
reverse_complement/simd_encoded/128
time: [25.035 ns 25.058 ns 25.086 ns]
thrpt: [4.7520 GiB/s 4.7573 GiB/s 4.7618 GiB/s]
change:
time: [−0.4240% −0.2751% −0.1266%] (p = 0.00 < 0.05)
thrpt: [+0.1268% +0.2759% +0.4258%]
Change within noise threshold.
Found 16 outliers among 100 measurements (16.00%)
1 (1.00%) low mild
7 (7.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/scalar/128
Benchmarking reverse_complement/scalar/128: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/128: Collecting 100 samples in estimated 5.0003 s (79M iterations)
Benchmarking reverse_complement/scalar/128: Analyzing
reverse_complement/scalar/128
time: [63.485 ns 63.560 ns 63.659 ns]
thrpt: [1.8726 GiB/s 1.8755 GiB/s 1.8778 GiB/s]
change:
time: [−1.0830% −0.4103% +0.0825%] (p = 0.20 > 0.05)
thrpt: [−0.0825% +0.4120% +1.0948%]
No change in performance detected.
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/255
Benchmarking reverse_complement/simd_high_level/255: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/255: Collecting 100 samples in estimated 5.0002 s (23M iterations)
Benchmarking reverse_complement/simd_high_level/255: Analyzing
reverse_complement/simd_high_level/255
time: [215.20 ns 217.56 ns 221.03 ns]
thrpt: [1.0744 GiB/s 1.0916 GiB/s 1.1035 GiB/s]
change:
time: [−0.8639% −0.0855% +0.7283%] (p = 0.84 > 0.05)
thrpt: [−0.7230% +0.0856% +0.8714%]
No change in performance detected.
Found 8 outliers among 100 measurements (8.00%)
3 (3.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_encoded/255
Benchmarking reverse_complement/simd_encoded/255: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/255: Collecting 100 samples in estimated 5.0002 s (84M iterations)
Benchmarking reverse_complement/simd_encoded/255: Analyzing
reverse_complement/simd_encoded/255
time: [59.548 ns 60.093 ns 60.908 ns]
thrpt: [3.8991 GiB/s 3.9520 GiB/s 3.9881 GiB/s]
change:
time: [+0.3600% +0.8109% +1.3939%] (p = 0.00 < 0.05)
thrpt: [−1.3747% −0.8043% −0.3588%]
Change within noise threshold.
Found 13 outliers among 100 measurements (13.00%)
5 (5.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/scalar/255
Benchmarking reverse_complement/scalar/255: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/255: Collecting 100 samples in estimated 5.0001 s (49M iterations)
Benchmarking reverse_complement/scalar/255: Analyzing
reverse_complement/scalar/255
time: [102.71 ns 102.77 ns 102.87 ns]
thrpt: [2.3086 GiB/s 2.3108 GiB/s 2.3123 GiB/s]
change:
time: [−0.2748% −0.1055% +0.0530%] (p = 0.22 > 0.05)
thrpt: [−0.0529% +0.1056% +0.2756%]
No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
5 (5.00%) high mild
2 (2.00%) high severe
Benchmarking reverse_complement/simd_high_level/256
Benchmarking reverse_complement/simd_high_level/256: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/256: Collecting 100 samples in estimated 5.0008 s (27M iterations)
Benchmarking reverse_complement/simd_high_level/256: Analyzing
reverse_complement/simd_high_level/256
time: [181.90 ns 182.06 ns 182.25 ns]
thrpt: [1.3082 GiB/s 1.3096 GiB/s 1.3107 GiB/s]
change:
time: [−0.1891% −0.0419% +0.0997%] (p = 0.57 > 0.05)
thrpt: [−0.0996% +0.0420% +0.1895%]
No change in performance detected.
Found 6 outliers among 100 measurements (6.00%)
4 (4.00%) high mild
2 (2.00%) high severe
Benchmarking reverse_complement/simd_encoded/256
Benchmarking reverse_complement/simd_encoded/256: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/256: Collecting 100 samples in estimated 5.0001 s (157M iterations)
Benchmarking reverse_complement/simd_encoded/256: Analyzing
reverse_complement/simd_encoded/256
time: [31.728 ns 31.753 ns 31.784 ns]
thrpt: [7.5013 GiB/s 7.5086 GiB/s 7.5144 GiB/s]
change:
time: [−0.1134% +0.0231% +0.1559%] (p = 0.74 > 0.05)
thrpt: [−0.1557% −0.0231% +0.1136%]
No change in performance detected.
Found 9 outliers among 100 measurements (9.00%)
3 (3.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/scalar/256
Benchmarking reverse_complement/scalar/256: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/256: Collecting 100 samples in estimated 5.0002 s (49M iterations)
Benchmarking reverse_complement/scalar/256: Analyzing
reverse_complement/scalar/256
time: [101.91 ns 101.96 ns 102.02 ns]
thrpt: [2.3369 GiB/s 2.3383 GiB/s 2.3395 GiB/s]
change:
time: [−0.2255% −0.0234% +0.1651%] (p = 0.82 > 0.05)
thrpt: [−0.1648% +0.0234% +0.2260%]
No change in performance detected.
Found 14 outliers among 100 measurements (14.00%)
1 (1.00%) low mild
6 (6.00%) high mild
7 (7.00%) high severe
Benchmarking reverse_complement/simd_high_level/512
Benchmarking reverse_complement/simd_high_level/512: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/512: Collecting 100 samples in estimated 5.0006 s (17M iterations)
Benchmarking reverse_complement/simd_high_level/512: Analyzing
reverse_complement/simd_high_level/512
time: [296.05 ns 296.32 ns 296.66 ns]
thrpt: [1.6074 GiB/s 1.6092 GiB/s 1.6107 GiB/s]
change:
time: [+0.1597% +0.3553% +0.5553%] (p = 0.00 < 0.05)
thrpt: [−0.5522% −0.3541% −0.1595%]
Change within noise threshold.
Found 14 outliers among 100 measurements (14.00%)
4 (4.00%) high mild
10 (10.00%) high severe
Benchmarking reverse_complement/simd_encoded/512
Benchmarking reverse_complement/simd_encoded/512: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/512: Collecting 100 samples in estimated 5.0000 s (125M iterations)
Benchmarking reverse_complement/simd_encoded/512: Analyzing
reverse_complement/simd_encoded/512
time: [39.678 ns 39.708 ns 39.749 ns]
thrpt: [11.996 GiB/s 12.009 GiB/s 12.018 GiB/s]
change:
time: [−0.1229% +0.0208% +0.1787%] (p = 0.79 > 0.05)
thrpt: [−0.1784% −0.0207% +0.1231%]
No change in performance detected.
Found 11 outliers among 100 measurements (11.00%)
1 (1.00%) low mild
3 (3.00%) high mild
7 (7.00%) high severe
Benchmarking reverse_complement/scalar/512
Benchmarking reverse_complement/scalar/512: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/512: Collecting 100 samples in estimated 5.0000 s (27M iterations)
Benchmarking reverse_complement/scalar/512: Analyzing
reverse_complement/scalar/512
time: [181.62 ns 181.72 ns 181.82 ns]
thrpt: [2.6225 GiB/s 2.6241 GiB/s 2.6254 GiB/s]
change:
time: [−0.1250% +0.0421% +0.2121%] (p = 0.64 > 0.05)
thrpt: [−0.2117% −0.0421% +0.1252%]
No change in performance detected.
Found 10 outliers among 100 measurements (10.00%)
5 (5.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/1023
Benchmarking reverse_complement/simd_high_level/1023: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/1023: Collecting 100 samples in estimated 5.0015 s (8.9M iterations)
Benchmarking reverse_complement/simd_high_level/1023: Analyzing
reverse_complement/simd_high_level/1023
time: [562.94 ns 567.46 ns 574.38 ns]
thrpt: [1.6587 GiB/s 1.6790 GiB/s 1.6924 GiB/s]
change:
time: [+0.0333% +0.4636% +1.0425%] (p = 0.07 > 0.05)
thrpt: [−1.0317% −0.4615% −0.0333%]
No change in performance detected.
Found 14 outliers among 100 measurements (14.00%)
6 (6.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/simd_encoded/1023
Benchmarking reverse_complement/simd_encoded/1023: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/1023: Collecting 100 samples in estimated 5.0002 s (61M iterations)
Benchmarking reverse_complement/simd_encoded/1023: Analyzing
reverse_complement/simd_encoded/1023
time: [81.852 ns 81.922 ns 82.016 ns]
thrpt: [11.617 GiB/s 11.630 GiB/s 11.640 GiB/s]
change:
time: [−0.6764% −0.2856% +0.0516%] (p = 0.13 > 0.05)
thrpt: [−0.0516% +0.2864% +0.6810%]
No change in performance detected.
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/scalar/1023
Benchmarking reverse_complement/scalar/1023: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/1023: Collecting 100 samples in estimated 5.0012 s (15M iterations)
Benchmarking reverse_complement/scalar/1023: Analyzing
reverse_complement/scalar/1023
time: [339.36 ns 339.59 ns 339.88 ns]
thrpt: [2.8031 GiB/s 2.8056 GiB/s 2.8075 GiB/s]
change:
time: [−0.7051% −0.3842% −0.0930%] (p = 0.02 < 0.05)
thrpt: [+0.0931% +0.3857% +0.7101%]
Change within noise threshold.
Found 7 outliers among 100 measurements (7.00%)
2 (2.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/1024
Benchmarking reverse_complement/simd_high_level/1024: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/1024: Collecting 100 samples in estimated 5.0024 s (9.5M iterations)
Benchmarking reverse_complement/simd_high_level/1024: Analyzing
reverse_complement/simd_high_level/1024
time: [526.45 ns 526.83 ns 527.32 ns]
thrpt: [1.8085 GiB/s 1.8102 GiB/s 1.8115 GiB/s]
change:
time: [−0.1941% −0.0219% +0.1283%] (p = 0.80 > 0.05)
thrpt: [−0.1282% +0.0219% +0.1944%]
No change in performance detected.
Found 13 outliers among 100 measurements (13.00%)
7 (7.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/simd_encoded/1024
Benchmarking reverse_complement/simd_encoded/1024: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/1024: Collecting 100 samples in estimated 5.0002 s (94M iterations)
Benchmarking reverse_complement/simd_encoded/1024: Analyzing
reverse_complement/simd_encoded/1024
time: [52.655 ns 52.696 ns 52.750 ns]
thrpt: [18.079 GiB/s 18.098 GiB/s 18.112 GiB/s]
change:
time: [−0.8077% −0.3002% +0.0201%] (p = 0.18 > 0.05)
thrpt: [−0.0201% +0.3011% +0.8143%]
No change in performance detected.
Found 5 outliers among 100 measurements (5.00%)
2 (2.00%) high mild
3 (3.00%) high severe
Benchmarking reverse_complement/scalar/1024
Benchmarking reverse_complement/scalar/1024: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/1024: Collecting 100 samples in estimated 5.0014 s (15M iterations)
Benchmarking reverse_complement/scalar/1024: Analyzing
reverse_complement/scalar/1024
time: [338.88 ns 339.11 ns 339.39 ns]
thrpt: [2.8100 GiB/s 2.8123 GiB/s 2.8142 GiB/s]
change:
time: [−0.4005% −0.2123% −0.0246%] (p = 0.03 < 0.05)
thrpt: [+0.0246% +0.2128% +0.4021%]
Change within noise threshold.
Found 13 outliers among 100 measurements (13.00%)
1 (1.00%) low mild
6 (6.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/simd_high_level/2048
Benchmarking reverse_complement/simd_high_level/2048: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/2048: Collecting 100 samples in estimated 5.0035 s (5.0M iterations)
Benchmarking reverse_complement/simd_high_level/2048: Analyzing
reverse_complement/simd_high_level/2048
time: [1.0073 µs 1.0081 µs 1.0091 µs]
thrpt: [1.8901 GiB/s 1.8920 GiB/s 1.8934 GiB/s]
change:
time: [−0.0207% +0.1545% +0.3370%] (p = 0.10 > 0.05)
thrpt: [−0.3359% −0.1543% +0.0207%]
No change in performance detected.
Found 8 outliers among 100 measurements (8.00%)
3 (3.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_encoded/2048
Benchmarking reverse_complement/simd_encoded/2048: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/2048: Collecting 100 samples in estimated 5.0000 s (56M iterations)
Benchmarking reverse_complement/simd_encoded/2048: Analyzing
reverse_complement/simd_encoded/2048
time: [89.972 ns 90.021 ns 90.085 ns]
thrpt: [21.173 GiB/s 21.188 GiB/s 21.199 GiB/s]
change:
time: [−0.1382% +0.0810% +0.3037%] (p = 0.49 > 0.05)
thrpt: [−0.3028% −0.0810% +0.1383%]
No change in performance detected.
Found 13 outliers among 100 measurements (13.00%)
3 (3.00%) high mild
10 (10.00%) high severe
Benchmarking reverse_complement/scalar/2048
Benchmarking reverse_complement/scalar/2048: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/2048: Collecting 100 samples in estimated 5.0030 s (7.8M iterations)
Benchmarking reverse_complement/scalar/2048: Analyzing
reverse_complement/scalar/2048
time: [637.68 ns 641.69 ns 649.52 ns]
thrpt: [2.9365 GiB/s 2.9724 GiB/s 2.9911 GiB/s]
change:
time: [−0.0904% +0.3854% +1.1201%] (p = 0.32 > 0.05)
thrpt: [−1.1077% −0.3839% +0.0904%]
No change in performance detected.
Found 15 outliers among 100 measurements (15.00%)
7 (7.00%) high mild
8 (8.00%) high severe
Benchmarking reverse_complement/simd_high_level/4095
Benchmarking reverse_complement/simd_high_level/4095: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/4095: Collecting 100 samples in estimated 5.0030 s (2.5M iterations)
Benchmarking reverse_complement/simd_high_level/4095: Analyzing
reverse_complement/simd_high_level/4095
time: [2.0317 µs 2.0348 µs 2.0390 µs]
thrpt: [1.8704 GiB/s 1.8742 GiB/s 1.8772 GiB/s]
change:
time: [+0.0456% +0.3652% +0.6731%] (p = 0.03 < 0.05)
thrpt: [−0.6686% −0.3639% −0.0456%]
Change within noise threshold.
Found 19 outliers among 100 measurements (19.00%)
2 (2.00%) high mild
17 (17.00%) high severe
Benchmarking reverse_complement/simd_encoded/4095
Benchmarking reverse_complement/simd_encoded/4095: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/4095: Collecting 100 samples in estimated 5.0003 s (22M iterations)
Benchmarking reverse_complement/simd_encoded/4095: Analyzing
reverse_complement/simd_encoded/4095
time: [223.67 ns 223.79 ns 223.92 ns]
thrpt: [17.032 GiB/s 17.042 GiB/s 17.051 GiB/s]
change:
time: [+0.0139% +0.1792% +0.3743%] (p = 0.04 < 0.05)
thrpt: [−0.3729% −0.1789% −0.0139%]
Change within noise threshold.
Found 9 outliers among 100 measurements (9.00%)
5 (5.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/scalar/4095
Benchmarking reverse_complement/scalar/4095: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/4095: Collecting 100 samples in estimated 5.0015 s (4.0M iterations)
Benchmarking reverse_complement/scalar/4095: Analyzing
reverse_complement/scalar/4095
time: [1.2469 µs 1.2486 µs 1.2508 µs]
thrpt: [3.0491 GiB/s 3.0544 GiB/s 3.0586 GiB/s]
change:
time: [−0.3234% −0.0638% +0.1845%] (p = 0.63 > 0.05)
thrpt: [−0.1842% +0.0638% +0.3245%]
No change in performance detected.
Found 8 outliers among 100 measurements (8.00%)
3 (3.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/4096
Benchmarking reverse_complement/simd_high_level/4096: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/4096: Collecting 100 samples in estimated 5.0085 s (2.5M iterations)
Benchmarking reverse_complement/simd_high_level/4096: Analyzing
reverse_complement/simd_high_level/4096
time: [1.9734 µs 1.9799 µs 1.9909 µs]
thrpt: [1.9160 GiB/s 1.9267 GiB/s 1.9330 GiB/s]
change:
time: [−0.3065% −0.0016% +0.3379%] (p = 0.99 > 0.05)
thrpt: [−0.3368% +0.0016% +0.3074%]
No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
3 (3.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/simd_encoded/4096
Benchmarking reverse_complement/simd_encoded/4096: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/4096: Collecting 100 samples in estimated 5.0009 s (29M iterations)
Benchmarking reverse_complement/simd_encoded/4096: Analyzing
reverse_complement/simd_encoded/4096
time: [171.43 ns 171.55 ns 171.73 ns]
thrpt: [22.214 GiB/s 22.237 GiB/s 22.252 GiB/s]
change:
time: [−0.1378% +0.0007% +0.1411%] (p = 0.98 > 0.05)
thrpt: [−0.1409% −0.0007% +0.1380%]
No change in performance detected.
Found 12 outliers among 100 measurements (12.00%)
1 (1.00%) low mild
6 (6.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/scalar/4096
Benchmarking reverse_complement/scalar/4096: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/4096: Collecting 100 samples in estimated 5.0051 s (4.0M iterations)
Benchmarking reverse_complement/scalar/4096: Analyzing
reverse_complement/scalar/4096
time: [1.2460 µs 1.2469 µs 1.2480 µs]
thrpt: [3.0567 GiB/s 3.0594 GiB/s 3.0616 GiB/s]
change:
time: [−0.1440% +0.0096% +0.1646%] (p = 0.90 > 0.05)
thrpt: [−0.1643% −0.0096% +0.1442%]
No change in performance detected.
Found 9 outliers among 100 measurements (9.00%)
4 (4.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/8192
Benchmarking reverse_complement/simd_high_level/8192: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/8192: Collecting 100 samples in estimated 5.0093 s (1.3M iterations)
Benchmarking reverse_complement/simd_high_level/8192: Analyzing
reverse_complement/simd_high_level/8192
time: [3.8772 µs 3.8810 µs 3.8859 µs]
thrpt: [1.9633 GiB/s 1.9658 GiB/s 1.9678 GiB/s]
change:
time: [+0.2562% +0.4177% +0.5715%] (p = 0.00 < 0.05)
thrpt: [−0.5683% −0.4160% −0.2555%]
Change within noise threshold.
Found 7 outliers among 100 measurements (7.00%)
4 (4.00%) high mild
3 (3.00%) high severe
Benchmarking reverse_complement/simd_encoded/8192
Benchmarking reverse_complement/simd_encoded/8192: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/8192: Collecting 100 samples in estimated 5.0001 s (16M iterations)
Benchmarking reverse_complement/simd_encoded/8192: Analyzing
reverse_complement/simd_encoded/8192
time: [315.07 ns 318.12 ns 322.59 ns]
thrpt: [23.651 GiB/s 23.982 GiB/s 24.215 GiB/s]
change:
time: [−1.5424% −0.8871% −0.2596%] (p = 0.01 < 0.05)
thrpt: [+0.2603% +0.8951% +1.5666%]
Change within noise threshold.
Found 10 outliers among 100 measurements (10.00%)
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/scalar/8192
Benchmarking reverse_complement/scalar/8192: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/8192: Collecting 100 samples in estimated 5.0073 s (2.0M iterations)
Benchmarking reverse_complement/scalar/8192: Analyzing
reverse_complement/scalar/8192
time: [2.4696 µs 2.4718 µs 2.4745 µs]
thrpt: [3.0833 GiB/s 3.0865 GiB/s 3.0893 GiB/s]
change:
time: [−0.1580% −0.0107% +0.1396%] (p = 0.88 > 0.05)
thrpt: [−0.1394% +0.0107% +0.1582%]
No change in performance detected.
Found 10 outliers among 100 measurements (10.00%)
5 (5.00%) high mild
5 (5.00%) high severe
Benchmarking reverse_complement/simd_high_level/9999
Benchmarking reverse_complement/simd_high_level/9999: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/9999: Collecting 100 samples in estimated 5.0142 s (1.0M iterations)
Benchmarking reverse_complement/simd_high_level/9999: Analyzing
reverse_complement/simd_high_level/9999
time: [4.8118 µs 4.8152 µs 4.8192 µs]
thrpt: [1.9323 GiB/s 1.9339 GiB/s 1.9353 GiB/s]
change:
time: [−0.1254% +0.0364% +0.2069%] (p = 0.67 > 0.05)
thrpt: [−0.2065% −0.0364% +0.1256%]
No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
3 (3.00%) high mild
4 (4.00%) high severe
Benchmarking reverse_complement/simd_encoded/9999
Benchmarking reverse_complement/simd_encoded/9999: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/9999: Collecting 100 samples in estimated 5.0014 s (11M iterations)
Benchmarking reverse_complement/simd_encoded/9999: Analyzing
reverse_complement/simd_encoded/9999
time: [465.52 ns 465.85 ns 466.27 ns]
thrpt: [19.972 GiB/s 19.990 GiB/s 20.004 GiB/s]
change:
time: [−0.0679% +0.0911% +0.2467%] (p = 0.26 > 0.05)
thrpt: [−0.2461% −0.0910% +0.0680%]
No change in performance detected.
Found 11 outliers among 100 measurements (11.00%)
5 (5.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/scalar/9999
Benchmarking reverse_complement/scalar/9999: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/9999: Collecting 100 samples in estimated 5.0010 s (1.7M iterations)
Benchmarking reverse_complement/scalar/9999: Analyzing
reverse_complement/scalar/9999
time: [3.0058 µs 3.0081 µs 3.0112 µs]
thrpt: [3.0926 GiB/s 3.0958 GiB/s 3.0981 GiB/s]
change:
time: [−0.5679% −0.3133% −0.0756%] (p = 0.01 < 0.05)
thrpt: [+0.0757% +0.3143% +0.5712%]
Change within noise threshold.
Found 14 outliers among 100 measurements (14.00%)
2 (2.00%) low mild
6 (6.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/simd_high_level/10000
Benchmarking reverse_complement/simd_high_level/10000: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_high_level/10000: Collecting 100 samples in estimated 5.0123 s (1.1M iterations)
Benchmarking reverse_complement/simd_high_level/10000: Analyzing
reverse_complement/simd_high_level/10000
time: [4.7373 µs 4.7423 µs 4.7481 µs]
thrpt: [1.9614 GiB/s 1.9639 GiB/s 1.9659 GiB/s]
change:
time: [−0.2354% −0.0511% +0.1321%] (p = 0.60 > 0.05)
thrpt: [−0.1320% +0.0511% +0.2360%]
No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
5 (5.00%) high mild
2 (2.00%) high severe
Benchmarking reverse_complement/simd_encoded/10000
Benchmarking reverse_complement/simd_encoded/10000: Warming up for 3.0000 s
Benchmarking reverse_complement/simd_encoded/10000: Collecting 100 samples in estimated 5.0011 s (13M iterations)
Benchmarking reverse_complement/simd_encoded/10000: Analyzing
reverse_complement/simd_encoded/10000
time: [388.48 ns 388.99 ns 389.68 ns]
thrpt: [23.900 GiB/s 23.942 GiB/s 23.974 GiB/s]
change:
time: [−0.1302% +0.0102% +0.1619%] (p = 0.89 > 0.05)
thrpt: [−0.1616% −0.0102% +0.1303%]
No change in performance detected.
Found 12 outliers among 100 measurements (12.00%)
2 (2.00%) low mild
4 (4.00%) high mild
6 (6.00%) high severe
Benchmarking reverse_complement/scalar/10000
Benchmarking reverse_complement/scalar/10000: Warming up for 3.0000 s
Benchmarking reverse_complement/scalar/10000: Collecting 100 samples in estimated 5.0021 s (1.6M iterations)
Benchmarking reverse_complement/scalar/10000: Analyzing
reverse_complement/scalar/10000
time: [3.0051 µs 3.0070 µs 3.0093 µs]
thrpt: [3.0949 GiB/s 3.0972 GiB/s 3.0991 GiB/s]
change:
time: [−0.6127% −0.1608% +0.1304%] (p = 0.53 > 0.05)
thrpt: [−0.1303% +0.1611% +0.6165%]
No change in performance detected.
Found 14 outliers among 100 measurements (14.00%)
2 (2.00%) low mild
7 (7.00%) high mild
5 (5.00%) high severe
