[libc-commits] [libc] [libc][math] Optimize maximum and minimum functions using builtins when available (PR #100002)
via libc-commits
libc-commits at lists.llvm.org
Mon Jul 22 16:20:44 PDT 2024
overmighty wrote:
Before:
<ul>
<li><details><summary>Intel Core i7-13700H, Clang 18</summary><ul>
<li><details><summary>fmaxf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 83718538 ns
Average runtime : 2.09296 ns/op
Ops per second : 477791907 op/s
-- Other function --
Total time : 55335013 ns
Average runtime : 1.38337 ns/op
Ops per second : 722870346 op/s
-- Average runtime ratio --
Mine / Other's : 1.51294
Performance tests with inputs in normal range:
-- My function --
Total time : 76354575 ns
Average runtime : 1.90886 ns/op
Ops per second : 523872210 op/s
-- Other function --
Total time : 52798788 ns
Average runtime : 1.31997 ns/op
Ops per second : 757593905 op/s
-- Average runtime ratio --
Mine / Other's : 1.44614
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 79979326 ns
Average runtime : 1.99948 ns/op
Ops per second : 500129746 op/s
-- Other function --
Total time : 53014929 ns
Average runtime : 1.32537 ns/op
Ops per second : 754505207 op/s
-- Average runtime ratio --
Mine / Other's : 1.50862 </code></pre></details></li>
<li><details><summary>fmaxf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2735968555 ns
Average runtime : 133.723 ns/op
Ops per second : 7478156 op/s
-- Other function --
Total time : 27224426 ns
Average runtime : 1.33062 ns/op
Ops per second : 751530996 op/s
-- Average runtime ratio --
Mine / Other's : 100.497
Performance tests with inputs in normal range:
-- My function --
Total time : 4417129285 ns
Average runtime : 7.18957 ns/op
Ops per second : 139090336 op/s
-- Other function --
Total time : 810857860 ns
Average runtime : 1.3198 ns/op
Ops per second : 757691366 op/s
-- Average runtime ratio --
Mine / Other's : 5.44748
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2945039405 ns
Average runtime : 7.19004 ns/op
Ops per second : 139081330 op/s
-- Other function --
Total time : 540534834 ns
Average runtime : 1.31967 ns/op
Ops per second : 757767999 op/s
-- Average runtime ratio --
Mine / Other's : 5.44838 </code></pre></details></li>
<li><details><summary>fmaximumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 57324820 ns
Average runtime : 1.43312 ns/op
Ops per second : 697778728 op/s
-- Other function --
Total time : 44271783 ns
Average runtime : 1.10679 ns/op
Ops per second : 903510933 op/s
-- Average runtime ratio --
Mine / Other's : 1.29484
Performance tests with inputs in normal range:
-- My function --
Total time : 53208916 ns
Average runtime : 1.33022 ns/op
Ops per second : 751754461 op/s
-- Other function --
Total time : 42239959 ns
Average runtime : 1.056 ns/op
Ops per second : 946971563 op/s
-- Average runtime ratio --
Mine / Other's : 1.25968
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 53294939 ns
Average runtime : 1.33237 ns/op
Ops per second : 750541059 op/s
-- Other function --
Total time : 42416670 ns
Average runtime : 1.06042 ns/op
Ops per second : 943026409 op/s
-- Average runtime ratio --
Mine / Other's : 1.25646 </code></pre></details></li>
<li><details><summary>fmaximumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2733990717 ns
Average runtime : 133.626 ns/op
Ops per second : 7483566 op/s
-- Other function --
Total time : 27182146 ns
Average runtime : 1.32855 ns/op
Ops per second : 752699952 op/s
-- Average runtime ratio --
Mine / Other's : 100.58
Performance tests with inputs in normal range:
-- My function --
Total time : 4407432956 ns
Average runtime : 7.17379 ns/op
Ops per second : 139396334 op/s
-- Other function --
Total time : 810801322 ns
Average runtime : 1.31971 ns/op
Ops per second : 757744201 op/s
-- Average runtime ratio --
Mine / Other's : 5.4359
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2937937609 ns
Average runtime : 7.1727 ns/op
Ops per second : 139417528 op/s
-- Other function --
Total time : 540654495 ns
Average runtime : 1.31996 ns/op
Ops per second : 757600285 op/s
-- Average runtime ratio --
Mine / Other's : 5.43404 </code></pre></details></li>
<li><details><summary>fmaximum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 121964961 ns
Average runtime : 3.04912 ns/op
Ops per second : 327963373 op/s
-- Other function --
Total time : 44268128 ns
Average runtime : 1.1067 ns/op
Ops per second : 903585532 op/s
-- Average runtime ratio --
Mine / Other's : 2.75514
Performance tests with inputs in normal range:
-- My function --
Total time : 116677300 ns
Average runtime : 2.91693 ns/op
Ops per second : 342826239 op/s
-- Other function --
Total time : 42229583 ns
Average runtime : 1.05574 ns/op
Ops per second : 947204238 op/s
-- Average runtime ratio --
Mine / Other's : 2.76293
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 116647999 ns
Average runtime : 2.9162 ns/op
Ops per second : 342912354 op/s
-- Other function --
Total time : 42420295 ns
Average runtime : 1.06051 ns/op
Ops per second : 942945823 op/s
-- Average runtime ratio --
Mine / Other's : 2.74982 </code></pre></details></li>
<li><details><summary>fmaximum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2736311797 ns
Average runtime : 133.74 ns/op
Ops per second : 7477218 op/s
-- Other function --
Total time : 27192683 ns
Average runtime : 1.32907 ns/op
Ops per second : 752408285 op/s
-- Average runtime ratio --
Mine / Other's : 100.627
Performance tests with inputs in normal range:
-- My function --
Total time : 5159571126 ns
Average runtime : 8.39801 ns/op
Ops per second : 119075788 op/s
-- Other function --
Total time : 810784210 ns
Average runtime : 1.31968 ns/op
Ops per second : 757760193 op/s
-- Average runtime ratio --
Mine / Other's : 6.36368
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 3439441856 ns
Average runtime : 8.39707 ns/op
Ops per second : 119089089 op/s
-- Other function --
Total time : 540653092 ns
Average runtime : 1.31995 ns/op
Ops per second : 757602251 op/s
-- Average runtime ratio --
Mine / Other's : 6.36164 </code></pre></details></li>
<li><details><summary>fminf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 56094325 ns
Average runtime : 1.40236 ns/op
Ops per second : 713085325 op/s
-- Other function --
Total time : 55334516 ns
Average runtime : 1.38336 ns/op
Ops per second : 722876838 op/s
-- Average runtime ratio --
Mine / Other's : 1.01373
Performance tests with inputs in normal range:
-- My function --
Total time : 52865515 ns
Average runtime : 1.32164 ns/op
Ops per second : 756637668 op/s
-- Other function --
Total time : 52789647 ns
Average runtime : 1.31974 ns/op
Ops per second : 757725089 op/s
-- Average runtime ratio --
Mine / Other's : 1.00144
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 53176228 ns
Average runtime : 1.3294 ns/op
Ops per second : 752216573 op/s
-- Other function --
Total time : 53016883 ns
Average runtime : 1.32542 ns/op
Ops per second : 754477399 op/s
-- Average runtime ratio --
Mine / Other's : 1.00301 </code></pre></details></li>
<li><details><summary>fminf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2723243474 ns
Average runtime : 133.101 ns/op
Ops per second : 7513099 op/s
-- Other function --
Total time : 27191011 ns
Average runtime : 1.32898 ns/op
Ops per second : 752454551 op/s
-- Average runtime ratio --
Mine / Other's : 100.152
Performance tests with inputs in normal range:
-- My function --
Total time : 5022600442 ns
Average runtime : 8.17507 ns/op
Ops per second : 122323088 op/s
-- Other function --
Total time : 810754210 ns
Average runtime : 1.31963 ns/op
Ops per second : 757788232 op/s
-- Average runtime ratio --
Mine / Other's : 6.19497
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 3348971937 ns
Average runtime : 8.1762 ns/op
Ops per second : 122306190 op/s
-- Other function --
Total time : 540627041 ns
Average runtime : 1.31989 ns/op
Ops per second : 757638758 op/s
-- Average runtime ratio --
Mine / Other's : 6.19461 </code></pre></details></li>
<li><details><summary>fminimumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 56098920 ns
Average runtime : 1.40247 ns/op
Ops per second : 713026917 op/s
-- Other function --
Total time : 44268225 ns
Average runtime : 1.1067 ns/op
Ops per second : 903583552 op/s
-- Average runtime ratio --
Mine / Other's : 1.26725
Performance tests with inputs in normal range:
-- My function --
Total time : 53321833 ns
Average runtime : 1.33304 ns/op
Ops per second : 750162508 op/s
-- Other function --
Total time : 42236565 ns
Average runtime : 1.05591 ns/op
Ops per second : 947047658 op/s
-- Average runtime ratio --
Mine / Other's : 1.26246
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 53508597 ns
Average runtime : 1.33771 ns/op
Ops per second : 747544175 op/s
-- Other function --
Total time : 42414396 ns
Average runtime : 1.06036 ns/op
Ops per second : 943076968 op/s
-- Average runtime ratio --
Mine / Other's : 1.26157 </code></pre></details></li>
<li><details><summary>fminimumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2721416470 ns
Average runtime : 133.012 ns/op
Ops per second : 7518143 op/s
-- Other function --
Total time : 27184599 ns
Average runtime : 1.32867 ns/op
Ops per second : 752632032 op/s
-- Average runtime ratio --
Mine / Other's : 100.109
Performance tests with inputs in normal range:
-- My function --
Total time : 4692254866 ns
Average runtime : 7.63738 ns/op
Ops per second : 130934916 op/s
-- Other function --
Total time : 810817381 ns
Average runtime : 1.31973 ns/op
Ops per second : 757729193 op/s
-- Average runtime ratio --
Mine / Other's : 5.78707
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 3128050757 ns
Average runtime : 7.63684 ns/op
Ops per second : 130944166 op/s
-- Other function --
Total time : 540653959 ns
Average runtime : 1.31996 ns/op
Ops per second : 757601037 op/s
-- Average runtime ratio --
Mine / Other's : 5.78568 </code></pre></details></li>
<li><details><summary>fminimum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 101254020 ns
Average runtime : 2.53135 ns/op
Ops per second : 395046438 op/s
-- Other function --
Total time : 44268299 ns
Average runtime : 1.10671 ns/op
Ops per second : 903582041 op/s
-- Average runtime ratio --
Mine / Other's : 2.28728
Performance tests with inputs in normal range:
-- My function --
Total time : 94634137 ns
Average runtime : 2.36585 ns/op
Ops per second : 422680876 op/s
-- Other function --
Total time : 42230209 ns
Average runtime : 1.05575 ns/op
Ops per second : 947190197 op/s
-- Average runtime ratio --
Mine / Other's : 2.24091
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 96387771 ns
Average runtime : 2.40969 ns/op
Ops per second : 414990818 op/s
-- Other function --
Total time : 42410602 ns
Average runtime : 1.06026 ns/op
Ops per second : 943161334 op/s
-- Average runtime ratio --
Mine / Other's : 2.27273 </code></pre></details></li>
<li><details><summary>fminimum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2729719615 ns
Average runtime : 133.417 ns/op
Ops per second : 7495275 op/s
-- Other function --
Total time : 27175069 ns
Average runtime : 1.3282 ns/op
Ops per second : 752895972 op/s
-- Average runtime ratio --
Mine / Other's : 100.449
Performance tests with inputs in normal range:
-- My function --
Total time : 5146693030 ns
Average runtime : 8.37705 ns/op
Ops per second : 119373740 op/s
-- Other function --
Total time : 810838876 ns
Average runtime : 1.31977 ns/op
Ops per second : 757709106 op/s
-- Average runtime ratio --
Mine / Other's : 6.34737
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 3431230321 ns
Average runtime : 8.37703 ns/op
Ops per second : 119374090 op/s
-- Other function --
Total time : 540677767 ns
Average runtime : 1.32001 ns/op
Ops per second : 757567677 op/s
-- Average runtime ratio --
Mine / Other's : 6.34616 </code></pre></details></li>
</li></ul></details>
<li><details><summary>Intel Core i7-13700H, Clang 18, <code>-march=native</code></summary><ul>
<li><details><summary>fmaxf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 85613121 ns
Average runtime : 2.14033 ns/op
Ops per second : 467218570 op/s
-- Other function --
Total time : 66478069 ns
Average runtime : 1.66195 ns/op
Ops per second : 601702796 op/s
-- Average runtime ratio --
Mine / Other's : 1.28784
Performance tests with inputs in normal range:
-- My function --
Total time : 75541392 ns
Average runtime : 1.88853 ns/op
Ops per second : 529511555 op/s
-- Other function --
Total time : 63548350 ns
Average runtime : 1.58871 ns/op
Ops per second : 629442621 op/s
-- Average runtime ratio --
Mine / Other's : 1.18872
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 76380676 ns
Average runtime : 1.90951 ns/op
Ops per second : 523693191 op/s
-- Other function --
Total time : 63616920 ns
Average runtime : 1.59042 ns/op
Ops per second : 628764171 op/s
-- Average runtime ratio --
Mine / Other's : 1.20063 </code></pre></details></li>
<li><details><summary>fmaxf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 126233310 ns
Average runtime : 6.16976 ns/op
Ops per second : 162080832 op/s
-- Other function --
Total time : 21818663 ns
Average runtime : 1.06641 ns/op
Ops per second : 937729319 op/s
-- Average runtime ratio --
Mine / Other's : 5.78557
Performance tests with inputs in normal range:
-- My function --
Total time : 3794560046 ns
Average runtime : 6.17624 ns/op
Ops per second : 161910733 op/s
-- Other function --
Total time : 651657499 ns
Average runtime : 1.06067 ns/op
Ops per second : 942795871 op/s
-- Average runtime ratio --
Mine / Other's : 5.82294
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2529646972 ns
Average runtime : 6.1759 ns/op
Ops per second : 161919826 op/s
-- Other function --
Total time : 432591185 ns
Average runtime : 1.05613 ns/op
Ops per second : 946852395 op/s
-- Average runtime ratio --
Mine / Other's : 5.84766 </code></pre></details></li>
<li><details><summary>fmaximumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 66427041 ns
Average runtime : 1.66067 ns/op
Ops per second : 602165012 op/s
-- Other function --
Total time : 55342722 ns
Average runtime : 1.38357 ns/op
Ops per second : 722769653 op/s
-- Average runtime ratio --
Mine / Other's : 1.20029
Performance tests with inputs in normal range:
-- My function --
Total time : 63354984 ns
Average runtime : 1.58387 ns/op
Ops per second : 631363745 op/s
-- Other function --
Total time : 52788592 ns
Average runtime : 1.31971 ns/op
Ops per second : 757740232 op/s
-- Average runtime ratio --
Mine / Other's : 1.20016
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 63674404 ns
Average runtime : 1.59186 ns/op
Ops per second : 628196535 op/s
-- Other function --
Total time : 53023324 ns
Average runtime : 1.32558 ns/op
Ops per second : 754385749 op/s
-- Average runtime ratio --
Mine / Other's : 1.20088 </code></pre></details></li>
<li><details><summary>fmaximumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 126561992 ns
Average runtime : 6.18583 ns/op
Ops per second : 161659908 op/s
-- Other function --
Total time : 21802056 ns
Average runtime : 1.06559 ns/op
Ops per second : 938443603 op/s
-- Average runtime ratio --
Mine / Other's : 5.80505
Performance tests with inputs in normal range:
-- My function --
Total time : 3788555197 ns
Average runtime : 6.16647 ns/op
Ops per second : 162167361 op/s
-- Other function --
Total time : 650149387 ns
Average runtime : 1.05822 ns/op
Ops per second : 944982818 op/s
-- Average runtime ratio --
Mine / Other's : 5.82721
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2529488383 ns
Average runtime : 6.17551 ns/op
Ops per second : 161929978 op/s
-- Other function --
Total time : 433941540 ns
Average runtime : 1.05943 ns/op
Ops per second : 943905946 op/s
-- Average runtime ratio --
Mine / Other's : 5.8291 </code></pre></details></li>
<li><details><summary>fmaximum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 113246490 ns
Average runtime : 2.83116 ns/op
Ops per second : 353212183 op/s
-- Other function --
Total time : 55341436 ns
Average runtime : 1.38353 ns/op
Ops per second : 722786448 op/s
-- Average runtime ratio --
Mine / Other's : 2.04632
Performance tests with inputs in normal range:
-- My function --
Total time : 106019717 ns
Average runtime : 2.65049 ns/op
Ops per second : 377288688 op/s
-- Other function --
Total time : 52786819 ns
Average runtime : 1.31967 ns/op
Ops per second : 757765683 op/s
-- Average runtime ratio --
Mine / Other's : 2.00845
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 108348683 ns
Average runtime : 2.70871 ns/op
Ops per second : 369178829 op/s
-- Other function --
Total time : 53020773 ns
Average runtime : 1.32552 ns/op
Ops per second : 754422045 op/s
-- Average runtime ratio --
Mine / Other's : 2.04351 </code></pre></details></li>
<li><details><summary>fmaximum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 141136003 ns
Average runtime : 6.89814 ns/op
Ops per second : 144966553 op/s
-- Other function --
Total time : 21815069 ns
Average runtime : 1.06623 ns/op
Ops per second : 937883808 op/s
-- Average runtime ratio --
Mine / Other's : 6.46966
Performance tests with inputs in normal range:
-- My function --
Total time : 4238240833 ns
Average runtime : 6.8984 ns/op
Ops per second : 144961087 op/s
-- Other function --
Total time : 651513726 ns
Average runtime : 1.06044 ns/op
Ops per second : 943003923 op/s
-- Average runtime ratio --
Mine / Other's : 6.50522
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2825517337 ns
Average runtime : 6.89824 ns/op
Ops per second : 144964603 op/s
-- Other function --
Total time : 434004679 ns
Average runtime : 1.05958 ns/op
Ops per second : 943768626 op/s
-- Average runtime ratio --
Mine / Other's : 6.51034 </code></pre></details></li>
<li><details><summary>fminf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 66400287 ns
Average runtime : 1.66001 ns/op
Ops per second : 602407637 op/s
-- Other function --
Total time : 66401998 ns
Average runtime : 1.66005 ns/op
Ops per second : 602392114 op/s
-- Average runtime ratio --
Mine / Other's : 0.999974
Performance tests with inputs in normal range:
-- My function --
Total time : 63388208 ns
Average runtime : 1.5847 ns/op
Ops per second : 631032825 op/s
-- Other function --
Total time : 63353066 ns
Average runtime : 1.58383 ns/op
Ops per second : 631382859 op/s
-- Average runtime ratio --
Mine / Other's : 1.00055
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 63632460 ns
Average runtime : 1.59081 ns/op
Ops per second : 628610617 op/s
-- Other function --
Total time : 63666705 ns
Average runtime : 1.59167 ns/op
Ops per second : 628272501 op/s
-- Average runtime ratio --
Mine / Other's : 0.999462 </code></pre></details></li>
<li><details><summary>fminf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 124149450 ns
Average runtime : 6.06791 ns/op
Ops per second : 164801374 op/s
-- Other function --
Total time : 21814070 ns
Average runtime : 1.06618 ns/op
Ops per second : 937926760 op/s
-- Average runtime ratio --
Mine / Other's : 5.69126
Performance tests with inputs in normal range:
-- My function --
Total time : 3756600844 ns
Average runtime : 6.11446 ns/op
Ops per second : 163546787 op/s
-- Other function --
Total time : 650146268 ns
Average runtime : 1.05822 ns/op
Ops per second : 944987351 op/s
-- Average runtime ratio --
Mine / Other's : 5.77809
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2501929443 ns
Average runtime : 6.10823 ns/op
Ops per second : 163713649 op/s
-- Other function --
Total time : 433990992 ns
Average runtime : 1.05955 ns/op
Ops per second : 943798391 op/s
-- Average runtime ratio --
Mine / Other's : 5.76493 </code></pre></details></li>
<li><details><summary>fminimumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 66423732 ns
Average runtime : 1.66059 ns/op
Ops per second : 602195010 op/s
-- Other function --
Total time : 55332363 ns
Average runtime : 1.38331 ns/op
Ops per second : 722904966 op/s
-- Average runtime ratio --
Mine / Other's : 1.20045
Performance tests with inputs in normal range:
-- My function --
Total time : 63353568 ns
Average runtime : 1.58384 ns/op
Ops per second : 631377857 op/s
-- Other function --
Total time : 52799373 ns
Average runtime : 1.31998 ns/op
Ops per second : 757585511 op/s
-- Average runtime ratio --
Mine / Other's : 1.19989
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 63621892 ns
Average runtime : 1.59055 ns/op
Ops per second : 628715034 op/s
-- Other function --
Total time : 53022357 ns
Average runtime : 1.32556 ns/op
Ops per second : 754399507 op/s
-- Average runtime ratio --
Mine / Other's : 1.19991 </code></pre></details></li>
<li><details><summary>fminimumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 126106446 ns
Average runtime : 6.16356 ns/op
Ops per second : 162243887 op/s
-- Other function --
Total time : 21815087 ns
Average runtime : 1.06623 ns/op
Ops per second : 937883034 op/s
-- Average runtime ratio --
Mine / Other's : 5.7807
Performance tests with inputs in normal range:
-- My function --
Total time : 3770172618 ns
Average runtime : 6.13655 ns/op
Ops per second : 162958055 op/s
-- Other function --
Total time : 650098699 ns
Average runtime : 1.05814 ns/op
Ops per second : 945056498 op/s
-- Average runtime ratio --
Mine / Other's : 5.79938
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2521353332 ns
Average runtime : 6.15565 ns/op
Ops per second : 162452439 op/s
-- Other function --
Total time : 432509897 ns
Average runtime : 1.05593 ns/op
Ops per second : 947030352 op/s
-- Average runtime ratio --
Mine / Other's : 5.82959 </code></pre></details></li>
<li><details><summary>fminimum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 99610412 ns
Average runtime : 2.49026 ns/op
Ops per second : 401564848 op/s
-- Other function --
Total time : 55333165 ns
Average runtime : 1.38333 ns/op
Ops per second : 722894488 op/s
-- Average runtime ratio --
Mine / Other's : 1.80019
Performance tests with inputs in normal range:
-- My function --
Total time : 91811664 ns
Average runtime : 2.29529 ns/op
Ops per second : 435674926 op/s
-- Other function --
Total time : 52786293 ns
Average runtime : 1.31966 ns/op
Ops per second : 757773234 op/s
-- Average runtime ratio --
Mine / Other's : 1.73931
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 95423266 ns
Average runtime : 2.38558 ns/op
Ops per second : 419185400 op/s
-- Other function --
Total time : 53016065 ns
Average runtime : 1.3254 ns/op
Ops per second : 754489040 op/s
-- Average runtime ratio --
Mine / Other's : 1.79989 </code></pre></details></li>
<li><details><summary>fminimum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 140446034 ns
Average runtime : 6.86442 ns/op
Ops per second : 145678730 op/s
-- Other function --
Total time : 21814455 ns
Average runtime : 1.0662 ns/op
Ops per second : 937910206 op/s
-- Average runtime ratio --
Mine / Other's : 6.43821
Performance tests with inputs in normal range:
-- My function --
Total time : 4236272764 ns
Average runtime : 6.8952 ns/op
Ops per second : 145028432 op/s
-- Other function --
Total time : 648699849 ns
Average runtime : 1.05586 ns/op
Ops per second : 947094408 op/s
-- Average runtime ratio --
Mine / Other's : 6.53041
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2823455539 ns
Average runtime : 6.8932 ns/op
Ops per second : 145070462 op/s
-- Other function --
Total time : 432519832 ns
Average runtime : 1.05596 ns/op
Ops per second : 947008598 op/s
-- Average runtime ratio --
Mine / Other's : 6.52792 </code></pre></details></li>
</li></ul></details>
<li><details><summary>Intel Core i7-13700H, GCC 14</summary><ul>
<li><details><summary>fmaxf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 72535954 ns
Average runtime : 1.8134 ns/op
Ops per second : 551451215 op/s
-- Other function --
Total time : 66469859 ns
Average runtime : 1.66174 ns/op
Ops per second : 601777115 op/s
-- Average runtime ratio --
Mine / Other's : 1.09126
Performance tests with inputs in normal range:
-- My function --
Total time : 69100940 ns
Average runtime : 1.72752 ns/op
Ops per second : 578863905 op/s
-- Other function --
Total time : 63298716 ns
Average runtime : 1.58247 ns/op
Ops per second : 631924982 op/s
-- Average runtime ratio --
Mine / Other's : 1.09166
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 84821803 ns
Average runtime : 2.12054 ns/op
Ops per second : 471577337 op/s
-- Other function --
Total time : 63522417 ns
Average runtime : 1.58806 ns/op
Ops per second : 629699590 op/s
-- Average runtime ratio --
Mine / Other's : 1.33531 </code></pre></details></li>
<li><details><summary>fmaxf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2719170890 ns
Average runtime : 132.902 ns/op
Ops per second : 7524352 op/s
-- Other function --
Total time : 26988516 ns
Average runtime : 1.31909 ns/op
Ops per second : 758100223 op/s
-- Average runtime ratio --
Mine / Other's : 100.753
Performance tests with inputs in normal range:
-- My function --
Total time : 4061137248 ns
Average runtime : 6.61014 ns/op
Ops per second : 151282747 op/s
-- Other function --
Total time : 704092798 ns
Average runtime : 1.14602 ns/op
Ops per second : 872583843 op/s
-- Average runtime ratio --
Mine / Other's : 5.7679
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2707755754 ns
Average runtime : 6.61073 ns/op
Ops per second : 151269182 op/s
-- Other function --
Total time : 461143950 ns
Average runtime : 1.12584 ns/op
Ops per second : 888225899 op/s
-- Average runtime ratio --
Mine / Other's : 5.87182 </code></pre></details></li>
<li><details><summary>fmaximumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 66684985 ns
Average runtime : 1.66712 ns/op
Ops per second : 599835780 op/s
-- Other function --
Total time : 55018775 ns
Average runtime : 1.37547 ns/op
Ops per second : 727025274 op/s
-- Average runtime ratio --
Mine / Other's : 1.21204
Performance tests with inputs in normal range:
-- My function --
Total time : 65952119 ns
Average runtime : 1.6488 ns/op
Ops per second : 606501210 op/s
-- Other function --
Total time : 52643277 ns
Average runtime : 1.31608 ns/op
Ops per second : 759831877 op/s
-- Average runtime ratio --
Mine / Other's : 1.25281
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 65176248 ns
Average runtime : 1.6294 ns/op
Ops per second : 613721121 op/s
-- Other function --
Total time : 53024372 ns
Average runtime : 1.32561 ns/op
Ops per second : 754370839 op/s
-- Average runtime ratio --
Mine / Other's : 1.22918 </code></pre></details></li>
<li><details><summary>fmaximumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2691777982 ns
Average runtime : 131.563 ns/op
Ops per second : 7600924 op/s
-- Other function --
Total time : 27052124 ns
Average runtime : 1.3222 ns/op
Ops per second : 756317692 op/s
-- Average runtime ratio --
Mine / Other's : 99.5034
Performance tests with inputs in normal range:
-- My function --
Total time : 4104703245 ns
Average runtime : 6.68105 ns/op
Ops per second : 149677080 op/s
-- Other function --
Total time : 706735543 ns
Average runtime : 1.15032 ns/op
Ops per second : 869320930 op/s
-- Average runtime ratio --
Mine / Other's : 5.80798
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2735682482 ns
Average runtime : 6.67891 ns/op
Ops per second : 149724978 op/s
-- Other function --
Total time : 466392235 ns
Average runtime : 1.13865 ns/op
Ops per second : 878230744 op/s
-- Average runtime ratio --
Mine / Other's : 5.86563 </code></pre></details></li>
<li><details><summary>fmaximum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 84692029 ns
Average runtime : 2.1173 ns/op
Ops per second : 472299937 op/s
-- Other function --
Total time : 55117889 ns
Average runtime : 1.37795 ns/op
Ops per second : 725717924 op/s
-- Average runtime ratio --
Mine / Other's : 1.53656
Performance tests with inputs in normal range:
-- My function --
Total time : 79849731 ns
Average runtime : 1.99624 ns/op
Ops per second : 500941449 op/s
-- Other function --
Total time : 51974535 ns
Average runtime : 1.29936 ns/op
Ops per second : 769608424 op/s
-- Average runtime ratio --
Mine / Other's : 1.53632
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 77794868 ns
Average runtime : 1.94487 ns/op
Ops per second : 514173248 op/s
-- Other function --
Total time : 52456483 ns
Average runtime : 1.31141 ns/op
Ops per second : 762537587 op/s
-- Average runtime ratio --
Mine / Other's : 1.48304 </code></pre></details></li>
<li><details><summary>fmaximum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2719806397 ns
Average runtime : 132.933 ns/op
Ops per second : 7522594 op/s
-- Other function --
Total time : 26350887 ns
Average runtime : 1.28792 ns/op
Ops per second : 776444451 op/s
-- Average runtime ratio --
Mine / Other's : 103.215
Performance tests with inputs in normal range:
-- My function --
Total time : 4233564583 ns
Average runtime : 6.89079 ns/op
Ops per second : 145121206 op/s
-- Other function --
Total time : 698458717 ns
Average runtime : 1.13685 ns/op
Ops per second : 879622495 op/s
-- Average runtime ratio --
Mine / Other's : 6.0613
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2822309476 ns
Average runtime : 6.8904 ns/op
Ops per second : 145129371 op/s
-- Other function --
Total time : 464264114 ns
Average runtime : 1.13346 ns/op
Ops per second : 882256430 op/s
-- Average runtime ratio --
Mine / Other's : 6.0791 </code></pre></details></li>
<li><details><summary>fminf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 78250500 ns
Average runtime : 1.95626 ns/op
Ops per second : 511179353 op/s
-- Other function --
Total time : 65480981 ns
Average runtime : 1.63702 ns/op
Ops per second : 610865008 op/s
-- Average runtime ratio --
Mine / Other's : 1.19501
Performance tests with inputs in normal range:
-- My function --
Total time : 65927035 ns
Average runtime : 1.64817 ns/op
Ops per second : 606731972 op/s
-- Other function --
Total time : 62243382 ns
Average runtime : 1.55608 ns/op
Ops per second : 642639244 op/s
-- Average runtime ratio --
Mine / Other's : 1.05918
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 68323626 ns
Average runtime : 1.70809 ns/op
Ops per second : 585449607 op/s
-- Other function --
Total time : 63364996 ns
Average runtime : 1.58412 ns/op
Ops per second : 631263986 op/s
-- Average runtime ratio --
Mine / Other's : 1.07826 </code></pre></details></li>
<li><details><summary>fminf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2705661061 ns
Average runtime : 132.241 ns/op
Ops per second : 7561922 op/s
-- Other function --
Total time : 27227960 ns
Average runtime : 1.33079 ns/op
Ops per second : 751433452 op/s
-- Average runtime ratio --
Mine / Other's : 99.3707
Performance tests with inputs in normal range:
-- My function --
Total time : 4105550069 ns
Average runtime : 6.68243 ns/op
Ops per second : 149646208 op/s
-- Other function --
Total time : 699710886 ns
Average runtime : 1.13889 ns/op
Ops per second : 878048365 op/s
-- Average runtime ratio --
Mine / Other's : 5.86749
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2737657603 ns
Average runtime : 6.68373 ns/op
Ops per second : 149616957 op/s
-- Other function --
Total time : 461921045 ns
Average runtime : 1.12774 ns/op
Ops per second : 886731627 op/s
-- Average runtime ratio --
Mine / Other's : 5.92668 </code></pre></details></li>
<li><details><summary>fminimumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 68204606 ns
Average runtime : 1.70511 ns/op
Ops per second : 586471242 op/s
-- Other function --
Total time : 55342239 ns
Average runtime : 1.38355 ns/op
Ops per second : 722775961 op/s
-- Average runtime ratio --
Mine / Other's : 1.23242
Performance tests with inputs in normal range:
-- My function --
Total time : 63970932 ns
Average runtime : 1.59927 ns/op
Ops per second : 625284621 op/s
-- Other function --
Total time : 52261625 ns
Average runtime : 1.30654 ns/op
Ops per second : 765380716 op/s
-- Average runtime ratio --
Mine / Other's : 1.22405
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 64351009 ns
Average runtime : 1.60877 ns/op
Ops per second : 621591496 op/s
-- Other function --
Total time : 52571480 ns
Average runtime : 1.31429 ns/op
Ops per second : 760869581 op/s
-- Average runtime ratio --
Mine / Other's : 1.22407 </code></pre></details></li>
<li><details><summary>fminimumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2718399821 ns
Average runtime : 132.864 ns/op
Ops per second : 7526486 op/s
-- Other function --
Total time : 27212139 ns
Average runtime : 1.33002 ns/op
Ops per second : 751870332 op/s
-- Average runtime ratio --
Mine / Other's : 99.8966
Performance tests with inputs in normal range:
-- My function --
Total time : 4060225809 ns
Average runtime : 6.60866 ns/op
Ops per second : 151316707 op/s
-- Other function --
Total time : 703916351 ns
Average runtime : 1.14573 ns/op
Ops per second : 872802569 op/s
-- Average runtime ratio --
Mine / Other's : 5.76805
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2707408294 ns
Average runtime : 6.60988 ns/op
Ops per second : 151288596 op/s
-- Other function --
Total time : 460398900 ns
Average runtime : 1.12402 ns/op
Ops per second : 889663289 op/s
-- Average runtime ratio --
Mine / Other's : 5.88057 </code></pre></details></li>
<li><details><summary>fminimum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 81806361 ns
Average runtime : 2.04516 ns/op
Ops per second : 488960021 op/s
-- Other function --
Total time : 55114561 ns
Average runtime : 1.37786 ns/op
Ops per second : 725761745 op/s
-- Average runtime ratio --
Mine / Other's : 1.4843
Performance tests with inputs in normal range:
-- My function --
Total time : 78283075 ns
Average runtime : 1.95707 ns/op
Ops per second : 510966642 op/s
-- Other function --
Total time : 52543904 ns
Average runtime : 1.3136 ns/op
Ops per second : 761268900 op/s
-- Average runtime ratio --
Mine / Other's : 1.48986
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 78153297 ns
Average runtime : 1.95383 ns/op
Ops per second : 511815131 op/s
-- Other function --
Total time : 52629849 ns
Average runtime : 1.31574 ns/op
Ops per second : 760025741 op/s
-- Average runtime ratio --
Mine / Other's : 1.48496 </code></pre></details></li>
<li><details><summary>fminimum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 2716695010 ns
Average runtime : 132.781 ns/op
Ops per second : 7531209 op/s
-- Other function --
Total time : 26337047 ns
Average runtime : 1.28725 ns/op
Ops per second : 776852469 op/s
-- Average runtime ratio --
Mine / Other's : 103.151
Performance tests with inputs in normal range:
-- My function --
Total time : 4232601517 ns
Average runtime : 6.88922 ns/op
Ops per second : 145154226 op/s
-- Other function --
Total time : 700089704 ns
Average runtime : 1.13951 ns/op
Ops per second : 877573254 op/s
-- Average runtime ratio --
Mine / Other's : 6.0458
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 2822032987 ns
Average runtime : 6.88973 ns/op
Ops per second : 145143590 op/s
-- Other function --
Total time : 466246376 ns
Average runtime : 1.1383 ns/op
Ops per second : 878505487 op/s
-- Average runtime ratio --
Mine / Other's : 6.05266 </code></pre></details></li>
</li></ul></details>
<li><details><summary>Intel Core i7-13700H, GCC 14, <code>-march=native</code></summary><ul>
<li><details><summary>fmaxf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 77268419 ns
Average runtime : 1.93171 ns/op
Ops per second : 517676439 op/s
-- Other function --
Total time : 55597068 ns
Average runtime : 1.38993 ns/op
Ops per second : 719463119 op/s
-- Average runtime ratio --
Mine / Other's : 1.38979
Performance tests with inputs in normal range:
-- My function --
Total time : 73901839 ns
Average runtime : 1.84754 ns/op
Ops per second : 541259061 op/s
-- Other function --
Total time : 53034983 ns
Average runtime : 1.32587 ns/op
Ops per second : 754219908 op/s
-- Average runtime ratio --
Mine / Other's : 1.39345
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 74257528 ns
Average runtime : 1.85644 ns/op
Ops per second : 538666463 op/s
-- Other function --
Total time : 53277266 ns
Average runtime : 1.33193 ns/op
Ops per second : 750790027 op/s
-- Average runtime ratio --
Mine / Other's : 1.39379 </code></pre></details></li>
<li><details><summary>fmaxf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 51652389 ns
Average runtime : 2.52455 ns/op
Ops per second : 396109461 op/s
-- Other function --
Total time : 21815418 ns
Average runtime : 1.06625 ns/op
Ops per second : 937868804 op/s
-- Average runtime ratio --
Mine / Other's : 2.3677
Performance tests with inputs in normal range:
-- My function --
Total time : 1548583185 ns
Average runtime : 2.52056 ns/op
Ops per second : 396736840 op/s
-- Other function --
Total time : 648741599 ns
Average runtime : 1.05593 ns/op
Ops per second : 947033458 op/s
-- Average runtime ratio --
Mine / Other's : 2.38706
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1032992643 ns
Average runtime : 2.52195 ns/op
Ops per second : 396517828 op/s
-- Other function --
Total time : 432660173 ns
Average runtime : 1.0563 ns/op
Ops per second : 946701419 op/s
-- Average runtime ratio --
Mine / Other's : 2.38754 </code></pre></details></li>
<li><details><summary>fmaximumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 67451397 ns
Average runtime : 1.68628 ns/op
Ops per second : 593020186 op/s
-- Other function --
Total time : 44272259 ns
Average runtime : 1.10681 ns/op
Ops per second : 903501219 op/s
-- Average runtime ratio --
Mine / Other's : 1.52356
Performance tests with inputs in normal range:
-- My function --
Total time : 63373093 ns
Average runtime : 1.58433 ns/op
Ops per second : 631183332 op/s
-- Other function --
Total time : 42231801 ns
Average runtime : 1.05579 ns/op
Ops per second : 947154491 op/s
-- Average runtime ratio --
Mine / Other's : 1.5006
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 64037916 ns
Average runtime : 1.60095 ns/op
Ops per second : 624630570 op/s
-- Other function --
Total time : 42419984 ns
Average runtime : 1.0605 ns/op
Ops per second : 942952736 op/s
-- Average runtime ratio --
Mine / Other's : 1.50962 </code></pre></details></li>
<li><details><summary>fmaximumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 41600666 ns
Average runtime : 2.03327 ns/op
Ops per second : 491819049 op/s
-- Other function --
Total time : 21806844 ns
Average runtime : 1.06583 ns/op
Ops per second : 938237555 op/s
-- Average runtime ratio --
Mine / Other's : 1.90769
Performance tests with inputs in normal range:
-- My function --
Total time : 1228807031 ns
Average runtime : 2.00008 ns/op
Ops per second : 499980863 op/s
-- Other function --
Total time : 648750389 ns
Average runtime : 1.05594 ns/op
Ops per second : 947020626 op/s
-- Average runtime ratio --
Mine / Other's : 1.89411
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 819493303 ns
Average runtime : 2.00072 ns/op
Ops per second : 499821046 op/s
-- Other function --
Total time : 432669666 ns
Average runtime : 1.05632 ns/op
Ops per second : 946680648 op/s
-- Average runtime ratio --
Mine / Other's : 1.89404 </code></pre></details></li>
<li><details><summary>fmaximum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 78702456 ns
Average runtime : 1.96756 ns/op
Ops per second : 508243859 op/s
-- Other function --
Total time : 44268938 ns
Average runtime : 1.10672 ns/op
Ops per second : 903568999 op/s
-- Average runtime ratio --
Mine / Other's : 1.77783
Performance tests with inputs in normal range:
-- My function --
Total time : 78560660 ns
Average runtime : 1.96401 ns/op
Ops per second : 509161201 op/s
-- Other function --
Total time : 42231133 ns
Average runtime : 1.05578 ns/op
Ops per second : 947169473 op/s
-- Average runtime ratio --
Mine / Other's : 1.86025
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 76992932 ns
Average runtime : 1.92482 ns/op
Ops per second : 519528727 op/s
-- Other function --
Total time : 42421610 ns
Average runtime : 1.06054 ns/op
Ops per second : 942916593 op/s
-- Average runtime ratio --
Mine / Other's : 1.81495 </code></pre></details></li>
<li><details><summary>fmaximum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 62376098 ns
Average runtime : 3.04869 ns/op
Ops per second : 328010258 op/s
-- Other function --
Total time : 21808997 ns
Average runtime : 1.06593 ns/op
Ops per second : 938144931 op/s
-- Average runtime ratio --
Mine / Other's : 2.86011
Performance tests with inputs in normal range:
-- My function --
Total time : 1845255044 ns
Average runtime : 3.00344 ns/op
Ops per second : 332951264 op/s
-- Other function --
Total time : 648761860 ns
Average runtime : 1.05596 ns/op
Ops per second : 947003882 op/s
-- Average runtime ratio --
Mine / Other's : 2.84427
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1230782447 ns
Average runtime : 3.00484 ns/op
Ops per second : 332796426 op/s
-- Other function --
Total time : 432607756 ns
Average runtime : 1.05617 ns/op
Ops per second : 946816126 op/s
-- Average runtime ratio --
Mine / Other's : 2.84503 </code></pre></details></li>
<li><details><summary>fminf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 55349992 ns
Average runtime : 1.38375 ns/op
Ops per second : 722674720 op/s
-- Other function --
Total time : 55585618 ns
Average runtime : 1.38964 ns/op
Ops per second : 719611321 op/s
-- Average runtime ratio --
Mine / Other's : 0.995761
Performance tests with inputs in normal range:
-- My function --
Total time : 52788312 ns
Average runtime : 1.31971 ns/op
Ops per second : 757744252 op/s
-- Other function --
Total time : 53057620 ns
Average runtime : 1.32644 ns/op
Ops per second : 753898120 op/s
-- Average runtime ratio --
Mine / Other's : 0.994924
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 53027264 ns
Average runtime : 1.32568 ns/op
Ops per second : 754329697 op/s
-- Other function --
Total time : 53279969 ns
Average runtime : 1.332 ns/op
Ops per second : 750751938 op/s
-- Average runtime ratio --
Mine / Other's : 0.995257 </code></pre></details></li>
<li><details><summary>fminf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 43844872 ns
Average runtime : 2.14296 ns/op
Ops per second : 466645221 op/s
-- Other function --
Total time : 21806237 ns
Average runtime : 1.0658 ns/op
Ops per second : 938263671 op/s
-- Average runtime ratio --
Mine / Other's : 2.01066
Performance tests with inputs in normal range:
-- My function --
Total time : 1309514367 ns
Average runtime : 2.13144 ns/op
Ops per second : 469166292 op/s
-- Other function --
Total time : 648810379 ns
Average runtime : 1.05604 ns/op
Ops per second : 946933063 op/s
-- Average runtime ratio --
Mine / Other's : 2.01833
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 871061584 ns
Average runtime : 2.12662 ns/op
Ops per second : 470230816 op/s
-- Other function --
Total time : 432650314 ns
Average runtime : 1.05628 ns/op
Ops per second : 946722992 op/s
-- Average runtime ratio --
Mine / Other's : 2.01332 </code></pre></details></li>
<li><details><summary>fminimumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 66489635 ns
Average runtime : 1.66224 ns/op
Ops per second : 601598128 op/s
-- Other function --
Total time : 44270726 ns
Average runtime : 1.10677 ns/op
Ops per second : 903532505 op/s
-- Average runtime ratio --
Mine / Other's : 1.50189
Performance tests with inputs in normal range:
-- My function --
Total time : 63347608 ns
Average runtime : 1.58369 ns/op
Ops per second : 631437259 op/s
-- Other function --
Total time : 42230276 ns
Average runtime : 1.05576 ns/op
Ops per second : 947188694 op/s
-- Average runtime ratio --
Mine / Other's : 1.50005
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 63805929 ns
Average runtime : 1.59515 ns/op
Ops per second : 626901615 op/s
-- Other function --
Total time : 42451639 ns
Average runtime : 1.06129 ns/op
Ops per second : 942249603 op/s
-- Average runtime ratio --
Mine / Other's : 1.50303 </code></pre></details></li>
<li><details><summary>fminimumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 51562637 ns
Average runtime : 2.52017 ns/op
Ops per second : 396798945 op/s
-- Other function --
Total time : 21799856 ns
Average runtime : 1.06549 ns/op
Ops per second : 938538309 op/s
-- Average runtime ratio --
Mine / Other's : 2.36527
Performance tests with inputs in normal range:
-- My function --
Total time : 1536799808 ns
Average runtime : 2.50138 ns/op
Ops per second : 399778811 op/s
-- Other function --
Total time : 648780269 ns
Average runtime : 1.05599 ns/op
Ops per second : 946977011 op/s
-- Average runtime ratio --
Mine / Other's : 2.36875
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1024118108 ns
Average runtime : 2.50029 ns/op
Ops per second : 399953869 op/s
-- Other function --
Total time : 432635089 ns
Average runtime : 1.05624 ns/op
Ops per second : 946756308 op/s
-- Average runtime ratio --
Mine / Other's : 2.36716 </code></pre></details></li>
<li><details><summary>fminimum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 97553132 ns
Average runtime : 2.43883 ns/op
Ops per second : 410033375 op/s
-- Other function --
Total time : 44268583 ns
Average runtime : 1.10671 ns/op
Ops per second : 903576245 op/s
-- Average runtime ratio --
Mine / Other's : 2.20367
Performance tests with inputs in normal range:
-- My function --
Total time : 92808158 ns
Average runtime : 2.3202 ns/op
Ops per second : 430997025 op/s
-- Other function --
Total time : 42233660 ns
Average runtime : 1.05584 ns/op
Ops per second : 947112800 op/s
-- Average runtime ratio --
Mine / Other's : 2.19749
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 87936228 ns
Average runtime : 2.1984 ns/op
Ops per second : 454875549 op/s
-- Other function --
Total time : 42411348 ns
Average runtime : 1.06028 ns/op
Ops per second : 943144745 op/s
-- Average runtime ratio --
Mine / Other's : 2.07341 </code></pre></details></li>
<li><details><summary>fminimum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 62178042 ns
Average runtime : 3.039 ns/op
Ops per second : 329055070 op/s
-- Other function --
Total time : 21819093 ns
Average runtime : 1.06643 ns/op
Ops per second : 937710838 op/s
-- Average runtime ratio --
Mine / Other's : 2.84971
Performance tests with inputs in normal range:
-- My function --
Total time : 1855436397 ns
Average runtime : 3.02001 ns/op
Ops per second : 331124257 op/s
-- Other function --
Total time : 648763117 ns
Average runtime : 1.05596 ns/op
Ops per second : 947002047 op/s
-- Average runtime ratio --
Mine / Other's : 2.85996
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1236637291 ns
Average runtime : 3.01913 ns/op
Ops per second : 331220805 op/s
-- Other function --
Total time : 432653132 ns
Average runtime : 1.05628 ns/op
Ops per second : 946716826 op/s
-- Average runtime ratio --
Mine / Other's : 2.85826 </code></pre></details></li>
</li></ul></details>
<li><details><summary>Google Tensor G3, Clang 17</summary><ul>
<li><details><summary>fmaxf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 75795736 ns
Average runtime : 1.89489 ns/op
Ops per second : 527734700 op/s
-- Other function --
Total time : 53211466 ns
Average runtime : 1.33029 ns/op
Ops per second : 751718436 op/s
-- Average runtime ratio --
Mine / Other's : 1.42442
Performance tests with inputs in normal range:
-- My function --
Total time : 71642904 ns
Average runtime : 1.79107 ns/op
Ops per second : 558325218 op/s
-- Other function --
Total time : 50812011 ns
Average runtime : 1.2703 ns/op
Ops per second : 787216235 op/s
-- Average runtime ratio --
Mine / Other's : 1.40996
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 71662272 ns
Average runtime : 1.79156 ns/op
Ops per second : 558174320 op/s
-- Other function --
Total time : 50985433 ns
Average runtime : 1.27463 ns/op
Ops per second : 784538595 op/s
-- Average runtime ratio --
Mine / Other's : 1.40554 </code></pre></details></li>
<li><details><summary>fmaxf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 28690429 ns
Average runtime : 1.40227 ns/op
Ops per second : 713129803 op/s
-- Other function --
Total time : 14577026 ns
Average runtime : 0.712465 ns/op
Ops per second : 1403578480 op/s
-- Average runtime ratio --
Mine / Other's : 1.96819
Performance tests with inputs in normal range:
-- My function --
Total time : 854737509 ns
Average runtime : 1.39122 ns/op
Ops per second : 718793774 op/s
-- Other function --
Total time : 429402588 ns
Average runtime : 0.69892 ns/op
Ops per second : 1430778521 op/s
-- Average runtime ratio --
Mine / Other's : 1.99053
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 567886027 ns
Average runtime : 1.38644 ns/op
Ops per second : 721271488 op/s
-- Other function --
Total time : 285304932 ns
Average runtime : 0.696545 ns/op
Ops per second : 1435656920 op/s
-- Average runtime ratio --
Mine / Other's : 1.99045 </code></pre></details></li>
<li><details><summary>fmaximumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 71507487 ns
Average runtime : 1.78769 ns/op
Ops per second : 559382544 op/s
-- Other function --
Total time : 36308390 ns
Average runtime : 0.907709 ns/op
Ops per second : 1101674847 op/s
-- Average runtime ratio --
Mine / Other's : 1.96945
Performance tests with inputs in normal range:
-- My function --
Total time : 68628418 ns
Average runtime : 1.71571 ns/op
Ops per second : 582849512 op/s
-- Other function --
Total time : 34836955 ns
Average runtime : 0.870923 ns/op
Ops per second : 1148207126 op/s
-- Average runtime ratio --
Mine / Other's : 1.96999
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 68630900 ns
Average runtime : 1.71577 ns/op
Ops per second : 582828434 op/s
-- Other function --
Total time : 34607503 ns
Average runtime : 0.865187 ns/op
Ops per second : 1155819881 op/s
-- Average runtime ratio --
Mine / Other's : 1.98312 </code></pre></details></li>
<li><details><summary>fmaximumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 43838094 ns
Average runtime : 2.14262 ns/op
Ops per second : 466717371 op/s
-- Other function --
Total time : 20248861 ns
Average runtime : 0.98968 ns/op
Ops per second : 1010427203 op/s
-- Average runtime ratio --
Mine / Other's : 2.16497
Performance tests with inputs in normal range:
-- My function --
Total time : 1293342978 ns
Average runtime : 2.10512 ns/op
Ops per second : 475032540 op/s
-- Other function --
Total time : 585760295 ns
Average runtime : 0.953417 ns/op
Ops per second : 1048859072 op/s
-- Average runtime ratio --
Mine / Other's : 2.20797
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 861770102 ns
Average runtime : 2.10393 ns/op
Ops per second : 475300778 op/s
-- Other function --
Total time : 390948486 ns
Average runtime : 0.954464 ns/op
Ops per second : 1047708367 op/s
-- Average runtime ratio --
Mine / Other's : 2.20431 </code></pre></details></li>
<li><details><summary>fmaximum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 143129435 ns
Average runtime : 3.57823 ns/op
Ops per second : 279467602 op/s
-- Other function --
Total time : 36526449 ns
Average runtime : 0.91316 ns/op
Ops per second : 1095097965 op/s
-- Average runtime ratio --
Mine / Other's : 3.91851
Performance tests with inputs in normal range:
-- My function --
Total time : 135592652 ns
Average runtime : 3.38981 ns/op
Ops per second : 295001531 op/s
-- Other function --
Total time : 34616496 ns
Average runtime : 0.865412 ns/op
Ops per second : 1155519611 op/s
-- Average runtime ratio --
Mine / Other's : 3.917
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 135953776 ns
Average runtime : 3.39884 ns/op
Ops per second : 294217940 op/s
-- Other function --
Total time : 35718872 ns
Average runtime : 0.892971 ns/op
Ops per second : 1119857312 op/s
-- Average runtime ratio --
Mine / Other's : 3.80622 </code></pre></details></li>
<li><details><summary>fmaximum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 75487915 ns
Average runtime : 3.68954 ns/op
Ops per second : 271036761 op/s
-- Other function --
Total time : 19719930 ns
Average runtime : 0.963828 ns/op
Ops per second : 1037529037 op/s
-- Average runtime ratio --
Mine / Other's : 3.828
Performance tests with inputs in normal range:
-- My function --
Total time : 2259568320 ns
Average runtime : 3.6778 ns/op
Ops per second : 271901493 op/s
-- Other function --
Total time : 585253744 ns
Average runtime : 0.952592 ns/op
Ops per second : 1049766885 op/s
-- Average runtime ratio --
Mine / Other's : 3.86084
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1504050660 ns
Average runtime : 3.672 ns/op
Ops per second : 272331252 op/s
-- Other function --
Total time : 389526612 ns
Average runtime : 0.950993 ns/op
Ops per second : 1051532776 op/s
-- Average runtime ratio --
Mine / Other's : 3.86123 </code></pre></details></li>
<li><details><summary>fminf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 106516642 ns
Average runtime : 2.66291 ns/op
Ops per second : 375528548 op/s
-- Other function --
Total time : 70983805 ns
Average runtime : 1.77459 ns/op
Ops per second : 563509380 op/s
-- Average runtime ratio --
Mine / Other's : 1.50058
Performance tests with inputs in normal range:
-- My function --
Total time : 102739462 ns
Average runtime : 2.56848 ns/op
Ops per second : 389334723 op/s
-- Other function --
Total time : 68058675 ns
Average runtime : 1.70147 ns/op
Ops per second : 587728750 op/s
-- Average runtime ratio --
Mine / Other's : 1.50957
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 102992839 ns
Average runtime : 2.57482 ns/op
Ops per second : 388376904 op/s
-- Other function --
Total time : 68036336 ns
Average runtime : 1.70091 ns/op
Ops per second : 587921724 op/s
-- Average runtime ratio --
Mine / Other's : 1.51379 </code></pre></details></li>
<li><details><summary>fminf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 28629272 ns
Average runtime : 1.39928 ns/op
Ops per second : 714653170 op/s
-- Other function --
Total time : 14585734 ns
Average runtime : 0.71289 ns/op
Ops per second : 1402740513 op/s
-- Average runtime ratio --
Mine / Other's : 1.96283
Performance tests with inputs in normal range:
-- My function --
Total time : 863295126 ns
Average runtime : 1.40515 ns/op
Ops per second : 711668560 op/s
-- Other function --
Total time : 429140178 ns
Average runtime : 0.698493 ns/op
Ops per second : 1431653411 op/s
-- Average runtime ratio --
Mine / Other's : 2.01169
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 574792196 ns
Average runtime : 1.4033 ns/op
Ops per second : 712605360 op/s
-- Other function --
Total time : 284787801 ns
Average runtime : 0.695283 ns/op
Ops per second : 1438263853 op/s
-- Average runtime ratio --
Mine / Other's : 2.01832 </code></pre></details></li>
<li><details><summary>fminimumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 105351603 ns
Average runtime : 2.63379 ns/op
Ops per second : 379681360 op/s
-- Other function --
Total time : 36227499 ns
Average runtime : 0.905687 ns/op
Ops per second : 1104134734 op/s
-- Average runtime ratio --
Mine / Other's : 2.90806
Performance tests with inputs in normal range:
-- My function --
Total time : 101117147 ns
Average runtime : 2.52793 ns/op
Ops per second : 395581176 op/s
-- Other function --
Total time : 34601441 ns
Average runtime : 0.865035 ns/op
Ops per second : 1156022374 op/s
-- Average runtime ratio --
Mine / Other's : 2.92234
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 102335653 ns
Average runtime : 2.55839 ns/op
Ops per second : 390871009 op/s
-- Other function --
Total time : 34811971 ns
Average runtime : 0.870298 ns/op
Ops per second : 1149031176 op/s
-- Average runtime ratio --
Mine / Other's : 2.93967 </code></pre></details></li>
<li><details><summary>fminimumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 43046142 ns
Average runtime : 2.10392 ns/op
Ops per second : 475303919 op/s
-- Other function --
Total time : 20696005 ns
Average runtime : 1.01153 ns/op
Ops per second : 988596591 op/s
-- Average runtime ratio --
Mine / Other's : 2.07993
Performance tests with inputs in normal range:
-- My function --
Total time : 1282778158 ns
Average runtime : 2.08792 ns/op
Ops per second : 478944855 op/s
-- Other function --
Total time : 587833944 ns
Average runtime : 0.956792 ns/op
Ops per second : 1045159107 op/s
-- Average runtime ratio --
Mine / Other's : 2.18221
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 857387248 ns
Average runtime : 2.09323 ns/op
Ops per second : 477730454 op/s
-- Other function --
Total time : 390570435 ns
Average runtime : 0.953541 ns/op
Ops per second : 1048722492 op/s
-- Average runtime ratio --
Mine / Other's : 2.19522 </code></pre></details></li>
<li><details><summary>fminimum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 148088135 ns
Average runtime : 3.7022 ns/op
Ops per second : 270109688 op/s
-- Other function --
Total time : 37533325 ns
Average runtime : 0.938332 ns/op
Ops per second : 1065720662 op/s
-- Average runtime ratio --
Mine / Other's : 3.94551
Performance tests with inputs in normal range:
-- My function --
Total time : 142208252 ns
Average runtime : 3.5552 ns/op
Ops per second : 281277910 op/s
-- Other function --
Total time : 34599690 ns
Average runtime : 0.864991 ns/op
Ops per second : 1156080878 op/s
-- Average runtime ratio --
Mine / Other's : 4.1101
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 141683024 ns
Average runtime : 3.54207 ns/op
Ops per second : 282320625 op/s
-- Other function --
Total time : 34740194 ns
Average runtime : 0.868504 ns/op
Ops per second : 1151405199 op/s
-- Average runtime ratio --
Mine / Other's : 4.07836 </code></pre></details></li>
<li><details><summary>fminimum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 75356568 ns
Average runtime : 3.68312 ns/op
Ops per second : 271509180 op/s
-- Other function --
Total time : 19697550 ns
Average runtime : 0.962735 ns/op
Ops per second : 1038707859 op/s
-- Average runtime ratio --
Mine / Other's : 3.82568
Performance tests with inputs in normal range:
-- My function --
Total time : 2259758627 ns
Average runtime : 3.67811 ns/op
Ops per second : 271878594 op/s
-- Other function --
Total time : 586750285 ns
Average runtime : 0.955028 ns/op
Ops per second : 1047089393 op/s
-- Average runtime ratio --
Mine / Other's : 3.85131
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1505676067 ns
Average runtime : 3.67597 ns/op
Ops per second : 272037265 op/s
-- Other function --
Total time : 390355835 ns
Average runtime : 0.953017 ns/op
Ops per second : 1049299032 op/s
-- Average runtime ratio --
Mine / Other's : 3.85719 </code></pre></details></li>
</li></ul></details>
<li><details><summary>Google Tensor G3, Clang 17, <code>-mcpu=cortex-x3</code></summary><ul>
<li><details><summary>fmaxf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 78270671 ns
Average runtime : 1.95676 ns/op
Ops per second : 511047618 op/s
-- Other function --
Total time : 53211060 ns
Average runtime : 1.33028 ns/op
Ops per second : 751724171 op/s
-- Average runtime ratio --
Mine / Other's : 1.47095
Performance tests with inputs in normal range:
-- My function --
Total time : 74500040 ns
Average runtime : 1.8625 ns/op
Ops per second : 536913000 op/s
-- Other function --
Total time : 50771281 ns
Average runtime : 1.26928 ns/op
Ops per second : 787847759 op/s
-- Average runtime ratio --
Mine / Other's : 1.46737
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 75172973 ns
Average runtime : 1.87932 ns/op
Ops per second : 532106665 op/s
-- Other function --
Total time : 51014486 ns
Average runtime : 1.27536 ns/op
Ops per second : 784091796 op/s
-- Average runtime ratio --
Mine / Other's : 1.47356 </code></pre></details></li>
<li><details><summary>fmaxf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 29380737 ns
Average runtime : 1.43601 ns/op
Ops per second : 696374634 op/s
-- Other function --
Total time : 18011515 ns
Average runtime : 0.880328 ns/op
Ops per second : 1135939980 op/s
-- Average runtime ratio --
Mine / Other's : 1.63122
Performance tests with inputs in normal range:
-- My function --
Total time : 865450033 ns
Average runtime : 1.40866 ns/op
Ops per second : 709896558 op/s
-- Other function --
Total time : 457207520 ns
Average runtime : 0.744177 ns/op
Ops per second : 1343766174 op/s
-- Average runtime ratio --
Mine / Other's : 1.8929
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 582485515 ns
Average runtime : 1.42208 ns/op
Ops per second : 703193451 op/s
-- Other function --
Total time : 311772380 ns
Average runtime : 0.761163 ns/op
Ops per second : 1313778981 op/s
-- Average runtime ratio --
Mine / Other's : 1.8683 </code></pre></details></li>
<li><details><summary>fmaximumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 106986165 ns
Average runtime : 2.67465 ns/op
Ops per second : 373880491 op/s
-- Other function --
Total time : 53319621 ns
Average runtime : 1.33299 ns/op
Ops per second : 750193629 op/s
-- Average runtime ratio --
Mine / Other's : 2.00651
Performance tests with inputs in normal range:
-- My function --
Total time : 101548096 ns
Average runtime : 2.5387 ns/op
Ops per second : 393902412 op/s
-- Other function --
Total time : 50762289 ns
Average runtime : 1.26906 ns/op
Ops per second : 787987318 op/s
-- Average runtime ratio --
Mine / Other's : 2.00046
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 101959147 ns
Average runtime : 2.54898 ns/op
Ops per second : 392314384 op/s
-- Other function --
Total time : 50984416 ns
Average runtime : 1.27461 ns/op
Ops per second : 784554244 op/s
-- Average runtime ratio --
Mine / Other's : 1.99981 </code></pre></details></li>
<li><details><summary>fmaximumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 28572021 ns
Average runtime : 1.39648 ns/op
Ops per second : 716085151 op/s
-- Other function --
Total time : 14893839 ns
Average runtime : 0.727949 ns/op
Ops per second : 1373722382 op/s
-- Average runtime ratio --
Mine / Other's : 1.91838
Performance tests with inputs in normal range:
-- My function --
Total time : 863056844 ns
Average runtime : 1.40476 ns/op
Ops per second : 711865046 op/s
-- Other function --
Total time : 427456461 ns
Average runtime : 0.695753 ns/op
Ops per second : 1437292580 op/s
-- Average runtime ratio --
Mine / Other's : 2.01905
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 575060181 ns
Average runtime : 1.40396 ns/op
Ops per second : 712273277 op/s
-- Other function --
Total time : 282931031 ns
Average runtime : 0.69075 ns/op
Ops per second : 1447702638 op/s
-- Average runtime ratio --
Mine / Other's : 2.03251 </code></pre></details></li>
<li><details><summary>fmaximum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 170485881 ns
Average runtime : 4.26214 ns/op
Ops per second : 234623769 op/s
-- Other function --
Total time : 53218547 ns
Average runtime : 1.33046 ns/op
Ops per second : 751618416 op/s
-- Average runtime ratio --
Mine / Other's : 3.2035
Performance tests with inputs in normal range:
-- My function --
Total time : 162974406 ns
Average runtime : 4.07436 ns/op
Ops per second : 245437556 op/s
-- Other function --
Total time : 51007121 ns
Average runtime : 1.27518 ns/op
Ops per second : 784205013 op/s
-- Average runtime ratio --
Mine / Other's : 3.19513
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 163500488 ns
Average runtime : 4.08751 ns/op
Ops per second : 244647832 op/s
-- Other function --
Total time : 51041911 ns
Average runtime : 1.27605 ns/op
Ops per second : 783670501 op/s
-- Average runtime ratio --
Mine / Other's : 3.20326 </code></pre></details></li>
<li><details><summary>fmaximum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 59324096 ns
Average runtime : 2.89952 ns/op
Ops per second : 344885154 op/s
-- Other function --
Total time : 15210042 ns
Average runtime : 0.743404 ns/op
Ops per second : 1345163938 op/s
-- Average runtime ratio --
Mine / Other's : 3.90032
Performance tests with inputs in normal range:
-- My function --
Total time : 1761444256 ns
Average runtime : 2.86703 ns/op
Ops per second : 348793325 op/s
-- Other function --
Total time : 431560506 ns
Average runtime : 0.702433 ns/op
Ops per second : 1423624246 op/s
-- Average runtime ratio --
Mine / Other's : 4.08157
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1177347738 ns
Average runtime : 2.87438 ns/op
Ops per second : 347900613 op/s
-- Other function --
Total time : 288924601 ns
Average runtime : 0.705382 ns/op
Ops per second : 1417670903 op/s
-- Average runtime ratio --
Mine / Other's : 4.07493 </code></pre></details></li>
<li><details><summary>fminf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 106519450 ns
Average runtime : 2.66298 ns/op
Ops per second : 375518649 op/s
-- Other function --
Total time : 53234294 ns
Average runtime : 1.33086 ns/op
Ops per second : 751396083 op/s
-- Average runtime ratio --
Mine / Other's : 2.00096
Performance tests with inputs in normal range:
-- My function --
Total time : 101758707 ns
Average runtime : 2.54397 ns/op
Ops per second : 393087148 op/s
-- Other function --
Total time : 50890828 ns
Average runtime : 1.27227 ns/op
Ops per second : 785997036 op/s
-- Average runtime ratio --
Mine / Other's : 1.99955
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 101965739 ns
Average runtime : 2.54914 ns/op
Ops per second : 392289021 op/s
-- Other function --
Total time : 51046997 ns
Average runtime : 1.27617 ns/op
Ops per second : 783592421 op/s
-- Average runtime ratio --
Mine / Other's : 1.99749 </code></pre></details></li>
<li><details><summary>fminf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 28505738 ns
Average runtime : 1.39324 ns/op
Ops per second : 717750229 op/s
-- Other function --
Total time : 14643718 ns
Average runtime : 0.715724 ns/op
Ops per second : 1397186151 op/s
-- Average runtime ratio --
Mine / Other's : 1.94662
Performance tests with inputs in normal range:
-- My function --
Total time : 864937664 ns
Average runtime : 1.40782 ns/op
Ops per second : 710317084 op/s
-- Other function --
Total time : 426626913 ns
Average runtime : 0.694402 ns/op
Ops per second : 1440087301 op/s
-- Average runtime ratio --
Mine / Other's : 2.02739
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 576935425 ns
Average runtime : 1.40853 ns/op
Ops per second : 709958137 op/s
-- Other function --
Total time : 284343465 ns
Average runtime : 0.694198 ns/op
Ops per second : 1440511389 op/s
-- Average runtime ratio --
Mine / Other's : 2.02901 </code></pre></details></li>
<li><details><summary>fminimumf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 106549235 ns
Average runtime : 2.66373 ns/op
Ops per second : 375413676 op/s
-- Other function --
Total time : 53305664 ns
Average runtime : 1.33264 ns/op
Ops per second : 750390052 op/s
-- Average runtime ratio --
Mine / Other's : 1.99884
Performance tests with inputs in normal range:
-- My function --
Total time : 101931234 ns
Average runtime : 2.54828 ns/op
Ops per second : 392421816 op/s
-- Other function --
Total time : 50915283 ns
Average runtime : 1.27288 ns/op
Ops per second : 785619516 op/s
-- Average runtime ratio --
Mine / Other's : 2.00198
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 102131592 ns
Average runtime : 2.55329 ns/op
Ops per second : 391651977 op/s
-- Other function --
Total time : 51014405 ns
Average runtime : 1.27536 ns/op
Ops per second : 784093041 op/s
-- Average runtime ratio --
Mine / Other's : 2.00201 </code></pre></details></li>
<li><details><summary>fminimumf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 28836670 ns
Average runtime : 1.40942 ns/op
Ops per second : 709513269 op/s
-- Other function --
Total time : 14553711 ns
Average runtime : 0.711325 ns/op
Ops per second : 1405827008 op/s
-- Average runtime ratio --
Mine / Other's : 1.9814
Performance tests with inputs in normal range:
-- My function --
Total time : 871677979 ns
Average runtime : 1.41879 ns/op
Ops per second : 704824504 op/s
-- Other function --
Total time : 431341716 ns
Average runtime : 0.702076 ns/op
Ops per second : 1424346352 op/s
-- Average runtime ratio --
Mine / Other's : 2.02085
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 585330608 ns
Average runtime : 1.42903 ns/op
Ops per second : 699775467 op/s
-- Other function --
Total time : 287774292 ns
Average runtime : 0.702574 ns/op
Ops per second : 1423337703 op/s
-- Average runtime ratio --
Mine / Other's : 2.03399 </code></pre></details></li>
<li><details><summary>fminimum_numf</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 170389852 ns
Average runtime : 4.25974 ns/op
Ops per second : 234755999 op/s
-- Other function --
Total time : 53455566 ns
Average runtime : 1.33639 ns/op
Ops per second : 748285781 op/s
-- Average runtime ratio --
Mine / Other's : 3.1875
Performance tests with inputs in normal range:
-- My function --
Total time : 162788086 ns
Average runtime : 4.0697 ns/op
Ops per second : 245718473 op/s
-- Other function --
Total time : 50874186 ns
Average runtime : 1.27185 ns/op
Ops per second : 786254152 op/s
-- Average runtime ratio --
Mine / Other's : 3.19982
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 163259603 ns
Average runtime : 4.08149 ns/op
Ops per second : 245008803 op/s
-- Other function --
Total time : 50985107 ns
Average runtime : 1.27463 ns/op
Ops per second : 784543611 op/s
-- Average runtime ratio --
Mine / Other's : 3.2021 </code></pre></details></li>
<li><details><summary>fminimum_numf16</summary><pre><code> Performance tests with inputs in denormal range:
-- My function --
Total time : 57668132 ns
Average runtime : 2.81858 ns/op
Ops per second : 354788672 op/s
-- Other function --
Total time : 14778687 ns
Average runtime : 0.722321 ns/op
Ops per second : 1384426099 op/s
-- Average runtime ratio --
Mine / Other's : 3.90211
Performance tests with inputs in normal range:
-- My function --
Total time : 1939929688 ns
Average runtime : 3.15754 ns/op
Ops per second : 316702199 op/s
-- Other function --
Total time : 549771688 ns
Average runtime : 0.89484 ns/op
Ops per second : 1117518441 op/s
-- Average runtime ratio --
Mine / Other's : 3.52861
Performance tests with inputs in normal range with exponents close to each other:
-- My function --
Total time : 1387130453 ns
Average runtime : 3.38655 ns/op
Ops per second : 295285853 op/s
-- Other function --
Total time : 367092408 ns
Average runtime : 0.896222 ns/op
Ops per second : 1115795344 op/s
-- Average runtime ratio --
Mine / Other's : 3.7787 </code></pre></details></li>
</li></ul></details>
</ul>
https://github.com/llvm/llvm-project/pull/100002
More information about the libc-commits
mailing list