{"id":1117,"date":"2026-10-11T19:31:24","date_gmt":"2026-10-11T17:31:24","guid":{"rendered":"https:\/\/lewoniewski.info\/blog\/?p=1117"},"modified":"2026-10-11T21:04:41","modified_gmt":"2026-10-11T19:04:41","slug":"python-3-14-kontra-python-3-15-test-wydajnosci","status":"publish","type":"post","link":"https:\/\/lewoniewski.info\/blog\/2026\/python-3-14-kontra-python-3-15-test-wydajnosci\/","title":{"rendered":"Python 3.14 kontra Python 3.15 \u2013 test wydajno\u015bci"},"content":{"rendered":"<p>Por\u00f3wnano wydajno\u015b\u0107 <a href=\"https:\/\/www.python.org\/downloads\/release\/python-3148\/\" rel=\"noopener\" target=\"_blank\">Python 3.14.8<\/a> i <a href=\"https:\/\/www.python.org\/downloads\/release\/python-3150\/\" rel=\"noopener\" target=\"_blank\">Python 3.15.0<\/a> na komputerach z procesorem serii AMD Ryzen 7000 oraz procesorem Intel Core 13. generacji. Jeden z nich u\u017cywany jest w komputerach stacjonarnych, drugi \u2013 w laptopach lub komputerach mini PC. W obu przypadkach u\u017cyto 64-bitowych kompilacji Pythona dzia\u0142aj\u0105cych w systemie Windows 11. Testy przeprowadzono za pomoc\u0105 pakietu <a href=\"https:\/\/pypi.org\/project\/pyperformance\/\" rel=\"noopener\" target=\"_blank\">pyperformance 1.14.0<\/a> i narz\u0119dzia pomiarowego pyperf 2.10.0.<!--more--><\/p>\n<style>\n.benchtab-scroll{overflow-x:auto;margin:1.25em 0 1.75em;max-width:100%}\n.benchtab{width:100%;border-collapse:collapse;font-variant-numeric:tabular-nums;line-height:1.45}\n.benchtab caption{text-align:left;font-weight:bold;margin-bottom:.6em}\n.benchtab th,.benchtab td{padding:.6em .75em;border-bottom:1px solid #dce2e8;vertical-align:top}\n.benchtab thead th{background:#f2f5f8;text-align:right}\n.benchtab thead th:first-child{text-align:left}\n.benchtab tbody th{font-weight:normal}\n.benchtab .nameben{text-align:left;min-width:14em}\n.benchtab .wartn{text-align:right;white-space:nowrap}\n.benchtab tr:nth-child(even){background:#fafbfc}\n.ben1,.ben1s{color:#1455a0}\n.ben2,.ben2s{color:#a52b25}\n.ben1s,.ben2s{font-weight:bold}\n.benvar{color:#596775}\n.ben1b{background:#e0e8ff}\n.bench-note{font-size:.9em;color:#4d5965}\n.bench-summary{padding:1em 1.2em;background:#f0f5fb;border-left:4px solid #1455a0;margin:1.5em 0}\ncode{font-size:.9em;overflow-wrap:anywhere}\n@media(max-width:600px){.benchtab{font-size:.88em}.benchtab th,.benchtab td{padding:.5em}}\n<\/style>\n<div class=\"see-also\">\n<p>Zobacz te\u017c: <a href=\"https:\/\/lewoniewski.info\/blog\/2025\/python-314-kontra-313-312-311-310-test-wydajnosci-wideo\/\">Python 3.14 kontra 3.13 \/ 3.12 \/ 3.11 \/ 3.10 \u2013 test wydajno\u015bci (wideo)<\/a><\/p>\n<\/div>\n<h2>Interpretacja wynik\u00f3w<\/h2>\n<p>Ka\u017cda warto\u015b\u0107 czasu w tabelach jest \u015bredni\u0105 arytmetyczn\u0105 ze \u015brednich czas\u00f3w danego testu w powtarzanych sesjach pomiarowych. Ka\u017cda sesja ma tak\u0105 sam\u0105 wag\u0119. Wyniki kalibracji i pomiar\u00f3w &#8222;rozgrzewkowych&#8221; s\u0105 pomijane. Wsp\u00f3\u0142czynniki przyspieszenia oblicza si\u0119 na podstawie u\u015brednionych czas\u00f3w, a nie przez u\u015brednienie wcze\u015bniej wyznaczonych wsp\u00f3\u0142czynnik\u00f3w. W ten spos\u00f3b ka\u017cdy test otrzymuje jeden zbiorczy wynik.<\/p>\n<p><strong>Kr\u00f3tszy czas oznacza lepszy wynik.<\/strong> Punktem odniesienia jest Python 3.14. Wynik <strong>1,50\u00d7 szybciej<\/strong> oznacza, \u017ce Python 3.15 potrzebuje oko\u0142o dw\u00f3ch trzecich czasu wersji referencyjnej, czyli <strong>o 33,33% mniej<\/strong>. Wynik <strong>1,20\u00d7 wolniej<\/strong> oznacza <strong>czas d\u0142u\u017cszy o 20%<\/strong>. Kolumna \u201eZmiana czasu\u201d przedstawia procentow\u0105 zmian\u0119 wzgl\u0119dem Pythona 3.14. Warto\u015bci ujemne oznaczaj\u0105 kr\u00f3tszy czas wykonania.<\/p>\n<p>Wynik og\u00f3lny jest \u015bredni\u0105 geometryczn\u0105 wsp\u00f3\u0142czynnik\u00f3w przyspieszenia poszczeg\u00f3lnych test\u00f3w. Ka\u017cdy test dost\u0119pny dla obu wersji ma tak\u0105 sam\u0105 wag\u0119, niezale\u017cnie od czasu jego wykonania. Jest to podsumowanie zestawu test\u00f3w, a nie prognoza przyspieszenia ca\u0142ej aplikacji. Czasy zaokr\u0105glono dla czytelno\u015bci; wszystkie obliczenia wykorzystuj\u0105 warto\u015bci przed zaokr\u0105gleniem.<\/p>\n<p><span class=\"ben1\">Kolor niebieski oznacza kr\u00f3tszy czas wykonania<\/span>, a <span class=\"ben2\">czerwony &#8211; d\u0142u\u017cszy<\/span>. Przyspieszenia i spowolnienia o wsp\u00f3\u0142czynniku co najmniej 1,10\u00d7 wyr\u00f3\u017cniono pogrubieniem. Symbol <strong>\u2020<\/strong> oznacza wyniki, kt\u00f3re wykazywa\u0142y du\u017c\u0105 zmienno\u015b\u0107 lub zmienia\u0142y kierunek r\u00f3\u017cnicy mi\u0119dzy sesjami. Wielko\u015b\u0107 takich zmian oraz niewielkie r\u00f3\u017cnice wzgl\u0119dem wyniku referencyjnego wymagaj\u0105 ostro\u017cnej interpretacji. Tabele przedstawiaj\u0105 u\u015brednione obserwacje i nie potwierdzaj\u0105 istotno\u015bci statystycznej ka\u017cdej r\u00f3\u017cnicy.<\/p>\n<div class=\"benchtab-scroll\">\n<table class=\"benchtab\" id=\"pybench-overview\">\n<caption>Og\u00f3lna wydajno\u015b\u0107 Pythona 3.15 wzgl\u0119dem Pythona 3.14<\/caption>\n<thead>\n<tr>\n<th scope=\"col\">Komputer<\/th>\n<th scope=\"col\">Testy dost\u0119pne dla obu wersji<\/th>\n<th scope=\"col\">\u015arednia geometryczna przyspieszenia<\/th>\n<th scope=\"col\">Odpowiadaj\u0105ca zmiana czasu<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td class=\"nameben\">AMD Ryzen 9 7900<\/td>\n<td class=\"wartn\">122<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,451\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-31,10%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\">Intel Core i3-1315U<\/td>\n<td class=\"wartn\">118<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,451\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-31,10%<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p class=\"bench-note\">Jednakowy zapis 1,451\u00d7 w obu wierszach wynika z zaokr\u0105glenia. Obliczone warto\u015bci to oko\u0142o 1,451413092\u00d7 dla AMD i 1,451321030\u00d7 dla Intela. Zestawy obejmuj\u0105 odpowiednio 122 i 118 test\u00f3w, wi\u0119c nie s\u0105 identyczne. Taki wynik nie oznacza jednakowej korzy\u015bci dla wszystkich aplikacji ani przewagi kt\u00f3rego\u015b producenta procesor\u00f3w.<\/p>\n<div class=\"bench-summary\">\n<p><strong>Python 3.15 jest szybszy w wi\u0119kszo\u015bci badanych zada\u0144. \u015arednia geometryczna wsp\u00f3\u0142czynnik\u00f3w przyspieszenia wynosi oko\u0142o 1,45\u00d7 na obu komputerach.<\/strong> Najwi\u0119ksze korzy\u015bci wyst\u0119puj\u0105 w testach kodowania Ascii85, Base32 i Base85. Po wy\u0142\u0105czeniu tych sze\u015bciu wyj\u0105tkowo du\u017cych przyspiesze\u0144 pozosta\u0142a cz\u0119\u015b\u0107 zestawu uzyskuje wynik oko\u0142o 1,20\u00d7. Niekt\u00f3re testy Base16 i wyra\u017ce\u0144 regularnych wykazuj\u0105 spadek wydajno\u015bci, dlatego korzy\u015b\u0107 zale\u017cy od rodzaju wykonywanych operacji.<\/p>\n<\/div>\n<h2>Komputer stacjonarny z procesorem AMD Ryzen 9 7900<\/h2>\n<p>Komputer stacjonarny AMD jest wyposa\u017cony w procesor Ryzen 9 7900, pami\u0119\u0107 RAM DDR5 oraz dysk M.2 PCIe Gen4 NVMe.<\/p>\n<p>Na tym komputerze dost\u0119pne s\u0105 wyniki <strong>122 test\u00f3w dla obu wersji<\/strong>. \u015arednia geometryczna wsp\u00f3\u0142czynnik\u00f3w przyspieszenia Pythona 3.15 wynosi <strong>1,451\u00d7<\/strong>. W znormalizowanym podsumowaniu zestawu odpowiada to skr\u00f3ceniu czasu o oko\u0142o <strong>31,1%<\/strong>. Mediana wsp\u00f3\u0142czynnik\u00f3w przyspieszenia wynosi <strong>1,184\u00d7<\/strong>, co pokazuje, \u017ce niewielka liczba wyj\u0105tkowo du\u017cych przyspiesze\u0144 podnosi wynik og\u00f3lny.<\/p>\n<p>Najwi\u0119ksze przyspieszenia wyst\u0119puj\u0105 w testach <code>base32_large<\/code> (<span class=\"ben1s\">134,748\u00d7 szybciej<\/span>), <code>ascii85_large<\/code> (<span class=\"ben1s\">68,675\u00d7 szybciej<\/span>), <code>base85_large<\/code> (<span class=\"ben1s\">63,446\u00d7 szybciej<\/span>). Warianty tych test\u00f3w kodowania dla ma\u0142ych danych r\u00f3wnie\u017c uzyskuj\u0105 du\u017ce przyspieszenia.<\/p>\n<p>Najwi\u0119ksze wyra\u017ane spowolnienia w u\u015brednionych wynikach wyst\u0119puj\u0105 w testach <code>base16_large<\/code> (<span class=\"ben2s\">1,223\u00d7 wolniej<\/span>), <code>base16_small<\/code> (<span class=\"ben2s\">1,222\u00d7 wolniej<\/span>), <code>regex_effbot<\/code> (<span class=\"ben2s\">1,222\u00d7 wolniej<\/span>). Spadki te s\u0105 szczeg\u00f3lnie istotne dla aplikacji, kt\u00f3re cz\u0119sto wykonuj\u0105 te operacje.<\/p>\n<p>Wyniki AMD by\u0142y zasadniczo sp\u00f3jne w powtarzanych pomiarach. Czas uruchamiania bez modu\u0142u <code>site<\/code> pozostaje bliski wynikowi wersji referencyjnej, a test <code>logging_silent<\/code> obejmuje pomiar o du\u017cej zmienno\u015bci. Z tego wzgl\u0119du wi\u0119ksze znaczenie przy interpretacji maj\u0105 du\u017ce, powtarzalne przyspieszenia ni\u017c niewielkie r\u00f3\u017cnice lub dok\u0142adna wielko\u015b\u0107 zmian w tych dw\u00f3ch testach.<\/p>\n<div class=\"benchtab-scroll\">\n<table class=\"benchtab\" id=\"pybench-amd\">\n<caption>AMD Ryzen 9 7900 \u2014 komputer stacjonarny: u\u015brednione czasy wykonania<\/caption>\n<thead>\n<tr>\n<th scope=\"col\">Test<\/th>\n<th scope=\"col\">Python 3.14.8<\/th>\n<th scope=\"col\">Python 3.15.0<\/th>\n<th scope=\"col\">Zmiana czasu<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td class=\"nameben\"><code>2to3<\/code><\/td>\n<td class=\"wartn\">229,3 ms<\/td>\n<td class=\"wartn\">207,6 ms (<span class=\"ben1s\">1,105\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,47%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>ascii85_large<\/code><\/td>\n<td class=\"wartn\">797,9 ms<\/td>\n<td class=\"wartn\">11,62 ms (<span class=\"ben1s\">68,675\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-98,54%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>ascii85_small<\/code><\/td>\n<td class=\"wartn\">14,88 ms<\/td>\n<td class=\"wartn\">0,3875 ms (<span class=\"ben1s\">38,394\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-97,40%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_generators<\/code><\/td>\n<td class=\"wartn\">269,8 ms<\/td>\n<td class=\"wartn\">237,9 ms (<span class=\"ben1s\">1,134\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,82%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_cpu_io_mixed<\/code><\/td>\n<td class=\"wartn\">420,2 ms<\/td>\n<td class=\"wartn\">393,4 ms (<span class=\"ben1\">1,068\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,38%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_cpu_io_mixed_tg<\/code><\/td>\n<td class=\"wartn\">404,4 ms<\/td>\n<td class=\"wartn\">380,8 ms (<span class=\"ben1\">1,062\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-5,83%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager<\/code><\/td>\n<td class=\"wartn\">124,2 ms<\/td>\n<td class=\"wartn\">98,19 ms (<span class=\"ben1s\">1,265\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-20,94%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_cpu_io_mixed<\/code><\/td>\n<td class=\"wartn\">349 ms<\/td>\n<td class=\"wartn\">322,5 ms (<span class=\"ben1\">1,082\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-7,60%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_cpu_io_mixed_tg<\/code><\/td>\n<td class=\"wartn\">365,3 ms<\/td>\n<td class=\"wartn\">342,9 ms (<span class=\"ben1\">1,065\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,12%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_io<\/code><\/td>\n<td class=\"wartn\">652,8 ms<\/td>\n<td class=\"wartn\">588,9 ms (<span class=\"ben1s\">1,108\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,78%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_io_tg<\/code><\/td>\n<td class=\"wartn\">610,1 ms<\/td>\n<td class=\"wartn\">558,3 ms (<span class=\"ben1\">1,093\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-8,49%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_memoization<\/code><\/td>\n<td class=\"wartn\">190,5 ms<\/td>\n<td class=\"wartn\">170,7 ms (<span class=\"ben1s\">1,116\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,40%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_memoization_tg<\/code><\/td>\n<td class=\"wartn\">243,9 ms<\/td>\n<td class=\"wartn\">224 ms (<span class=\"ben1\">1,089\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-8,17%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_tg<\/code><\/td>\n<td class=\"wartn\">184,9 ms<\/td>\n<td class=\"wartn\">166,3 ms (<span class=\"ben1s\">1,112\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,07%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_io<\/code><\/td>\n<td class=\"wartn\">617,5 ms<\/td>\n<td class=\"wartn\">558,8 ms (<span class=\"ben1s\">1,105\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,50%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_io_tg<\/code><\/td>\n<td class=\"wartn\">581,5 ms<\/td>\n<td class=\"wartn\">511,6 ms (<span class=\"ben1s\">1,137\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,02%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_memoization<\/code><\/td>\n<td class=\"wartn\">308,5 ms<\/td>\n<td class=\"wartn\">278 ms (<span class=\"ben1s\">1,109\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,86%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_memoization_tg<\/code><\/td>\n<td class=\"wartn\">267,1 ms<\/td>\n<td class=\"wartn\">240,1 ms (<span class=\"ben1s\">1,112\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,11%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_none<\/code><\/td>\n<td class=\"wartn\">248,5 ms<\/td>\n<td class=\"wartn\">225,8 ms (<span class=\"ben1s\">1,101\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,15%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_none_tg<\/code><\/td>\n<td class=\"wartn\">215,4 ms<\/td>\n<td class=\"wartn\">194,4 ms (<span class=\"ben1s\">1,108\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,73%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio_tcp<\/code><\/td>\n<td class=\"wartn\">458,4 ms<\/td>\n<td class=\"wartn\">435,2 ms (<span class=\"ben1\">1,053\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-5,05%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio_tcp_ssl<\/code><\/td>\n<td class=\"wartn\">1,258 s<\/td>\n<td class=\"wartn\">1,219 s (<span class=\"ben1\">1,031\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-3,05%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio_websockets<\/code><\/td>\n<td class=\"wartn\">141,6 ms<\/td>\n<td class=\"wartn\">125,6 ms (<span class=\"ben1s\">1,128\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,33%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base16_large<\/code><\/td>\n<td class=\"wartn\">4,714 ms<\/td>\n<td class=\"wartn\">5,767 ms (<span class=\"ben2s\">1,223\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+22,34%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base16_small<\/code><\/td>\n<td class=\"wartn\">252,3 \u00b5s<\/td>\n<td class=\"wartn\">308,3 \u00b5s (<span class=\"ben2s\">1,222\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+22,20%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base32_large<\/code><\/td>\n<td class=\"wartn\">327,6 ms<\/td>\n<td class=\"wartn\">2,431 ms (<span class=\"ben1s\">134,748\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-99,26%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base32_small<\/code><\/td>\n<td class=\"wartn\">6,227 ms<\/td>\n<td class=\"wartn\">0,163 ms (<span class=\"ben1s\">38,200\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-97,38%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base64_large<\/code><\/td>\n<td class=\"wartn\">6,746 ms<\/td>\n<td class=\"wartn\">1,852 ms (<span class=\"ben1s\">3,643\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-72,55%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base64_small<\/code><\/td>\n<td class=\"wartn\">219,5 \u00b5s<\/td>\n<td class=\"wartn\">169,2 \u00b5s (<span class=\"ben1s\">1,297\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-22,91%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base85_large<\/code><\/td>\n<td class=\"wartn\">272,4 ms<\/td>\n<td class=\"wartn\">4,293 ms (<span class=\"ben1s\">63,446\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-98,42%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base85_small<\/code><\/td>\n<td class=\"wartn\">4,864 ms<\/td>\n<td class=\"wartn\">0,1505 ms (<span class=\"ben1s\">32,313\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-96,91%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>bench_mp_pool<\/code><\/td>\n<td class=\"wartn\">126 ms<\/td>\n<td class=\"wartn\">112 ms (<span class=\"ben1s\">1,125\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,08%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>bench_thread_pool<\/code><\/td>\n<td class=\"wartn\">738,9 \u00b5s<\/td>\n<td class=\"wartn\">658,8 \u00b5s (<span class=\"ben1s\">1,122\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,84%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>bpe_tokeniser<\/code><\/td>\n<td class=\"wartn\">2,968 s<\/td>\n<td class=\"wartn\">2,318 s (<span class=\"ben1s\">1,280\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-21,90%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>chameleon<\/code><\/td>\n<td class=\"wartn\">9,954 ms<\/td>\n<td class=\"wartn\">8,428 ms (<span class=\"ben1s\">1,181\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-15,33%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>chaos<\/code><\/td>\n<td class=\"wartn\">44,06 ms<\/td>\n<td class=\"wartn\">31,3 ms (<span class=\"ben1s\">1,408\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-28,98%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>comprehensions<\/code><\/td>\n<td class=\"wartn\">12,2 \u00b5s<\/td>\n<td class=\"wartn\">8,117 \u00b5s (<span class=\"ben1s\">1,503\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-33,47%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>connected_components<\/code><\/td>\n<td class=\"wartn\">388,3 ms<\/td>\n<td class=\"wartn\">365,8 ms (<span class=\"ben1\">1,061\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-5,78%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>coroutines<\/code><\/td>\n<td class=\"wartn\">17,05 ms<\/td>\n<td class=\"wartn\">12,87 ms (<span class=\"ben1s\">1,325\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-24,50%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>coverage<\/code><\/td>\n<td class=\"wartn\">52,53 ms<\/td>\n<td class=\"wartn\">48,64 ms (<span class=\"ben1\">1,080\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-7,39%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>create_gc_cycles<\/code><\/td>\n<td class=\"wartn\">1,03 ms<\/td>\n<td class=\"wartn\">0,8892 ms (<span class=\"ben1s\">1,158\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,66%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>crypto_pyaes<\/code><\/td>\n<td class=\"wartn\">50,48 ms<\/td>\n<td class=\"wartn\">39,17 ms (<span class=\"ben1s\">1,289\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-22,41%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>dask<\/code><\/td>\n<td class=\"wartn\">720,2 ms<\/td>\n<td class=\"wartn\">687,5 ms (<span class=\"ben1\">1,048\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-4,54%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deepcopy<\/code><\/td>\n<td class=\"wartn\">174,8 \u00b5s<\/td>\n<td class=\"wartn\">129,9 \u00b5s (<span class=\"ben1s\">1,346\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,69%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deepcopy_memo<\/code><\/td>\n<td class=\"wartn\">19,74 \u00b5s<\/td>\n<td class=\"wartn\">14,64 \u00b5s (<span class=\"ben1s\">1,349\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,86%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deepcopy_reduce<\/code><\/td>\n<td class=\"wartn\">1,899 \u00b5s<\/td>\n<td class=\"wartn\">1,541 \u00b5s (<span class=\"ben1s\">1,232\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-18,83%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deltablue<\/code><\/td>\n<td class=\"wartn\">2,508 ms<\/td>\n<td class=\"wartn\">1,669 ms (<span class=\"ben1s\">1,503\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-33,46%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>django_template<\/code><\/td>\n<td class=\"wartn\">24,56 ms<\/td>\n<td class=\"wartn\">20,47 ms (<span class=\"ben1s\">1,200\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,67%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>docutils<\/code><\/td>\n<td class=\"wartn\">1,425 s<\/td>\n<td class=\"wartn\">1,272 s (<span class=\"ben1s\">1,120\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,74%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>dulwich_log<\/code><\/td>\n<td class=\"wartn\">48,84 ms<\/td>\n<td class=\"wartn\">44,74 ms (<span class=\"ben1\">1,092\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-8,40%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>fannkuch<\/code><\/td>\n<td class=\"wartn\">280,5 ms<\/td>\n<td class=\"wartn\">208,5 ms (<span class=\"ben1s\">1,345\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,65%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>fastapi_http<\/code><\/td>\n<td class=\"wartn\">454,7 ms<\/td>\n<td class=\"wartn\">399,9 ms (<span class=\"ben1s\">1,137\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,06%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>float<\/code><\/td>\n<td class=\"wartn\">57,28 ms<\/td>\n<td class=\"wartn\">44,02 ms (<span class=\"ben1s\">1,301\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,16%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>gc_traversal<\/code><\/td>\n<td class=\"wartn\">1,807 ms<\/td>\n<td class=\"wartn\">1,682 ms (<span class=\"ben1\">1,074\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,89%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>generators<\/code><\/td>\n<td class=\"wartn\">23,32 ms<\/td>\n<td class=\"wartn\">16,29 ms (<span class=\"ben1s\">1,431\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-30,14%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>go<\/code><\/td>\n<td class=\"wartn\">89,44 ms<\/td>\n<td class=\"wartn\">59,34 ms (<span class=\"ben1s\">1,507\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-33,65%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>hexiom<\/code><\/td>\n<td class=\"wartn\">4,914 ms<\/td>\n<td class=\"wartn\">3,178 ms (<span class=\"ben1s\">1,546\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-35,32%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>html5lib<\/code><\/td>\n<td class=\"wartn\">33,73 ms<\/td>\n<td class=\"wartn\">28,44 ms (<span class=\"ben1s\">1,186\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-15,68%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>json_dumps<\/code><\/td>\n<td class=\"wartn\">6,731 ms<\/td>\n<td class=\"wartn\">4,862 ms (<span class=\"ben1s\">1,384\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-27,77%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>json_loads<\/code><\/td>\n<td class=\"wartn\">14,53 \u00b5s<\/td>\n<td class=\"wartn\">13,06 \u00b5s (<span class=\"ben1s\">1,113\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,15%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>k_core<\/code><\/td>\n<td class=\"wartn\">1,708 s<\/td>\n<td class=\"wartn\">1,616 s (<span class=\"ben1\">1,057\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-5,39%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>logging_format<\/code><\/td>\n<td class=\"wartn\">6,286 \u00b5s<\/td>\n<td class=\"wartn\">5,081 \u00b5s (<span class=\"ben1s\">1,237\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,18%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>logging_silent<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">75,05 ns<\/td>\n<td class=\"wartn\">51,26 ns (<span class=\"benvar\">1,464\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-31,69%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>logging_simple<\/code><\/td>\n<td class=\"wartn\">5,845 \u00b5s<\/td>\n<td class=\"wartn\">4,72 \u00b5s (<span class=\"ben1s\">1,238\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,25%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>mako<\/code><\/td>\n<td class=\"wartn\">7,97 ms<\/td>\n<td class=\"wartn\">6,643 ms (<span class=\"ben1s\">1,200\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,65%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>many_optionals<\/code><\/td>\n<td class=\"wartn\">386,7 \u00b5s<\/td>\n<td class=\"wartn\">297,7 \u00b5s (<span class=\"ben1s\">1,299\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,02%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>mdp<\/code><\/td>\n<td class=\"wartn\">829,3 ms<\/td>\n<td class=\"wartn\">661,3 ms (<span class=\"ben1s\">1,254\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-20,25%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>meteor_contest<\/code><\/td>\n<td class=\"wartn\">69,78 ms<\/td>\n<td class=\"wartn\">60,92 ms (<span class=\"ben1s\">1,145\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,70%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>nbody<\/code><\/td>\n<td class=\"wartn\">89,91 ms<\/td>\n<td class=\"wartn\">62,13 ms (<span class=\"ben1s\">1,447\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-30,89%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>nqueens<\/code><\/td>\n<td class=\"wartn\">65,02 ms<\/td>\n<td class=\"wartn\">43,38 ms (<span class=\"ben1s\">1,499\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-33,28%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pathlib<\/code><\/td>\n<td class=\"wartn\">232,9 ms<\/td>\n<td class=\"wartn\">228,7 ms (<span class=\"ben1\">1,018\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-1,80%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle<\/code><\/td>\n<td class=\"wartn\">7,953 \u00b5s<\/td>\n<td class=\"wartn\">7,807 \u00b5s (<span class=\"ben1\">1,019\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-1,83%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle_dict<\/code><\/td>\n<td class=\"wartn\">20,5 \u00b5s<\/td>\n<td class=\"wartn\">17,69 \u00b5s (<span class=\"ben1s\">1,159\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,73%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle_list<\/code><\/td>\n<td class=\"wartn\">3,041 \u00b5s<\/td>\n<td class=\"wartn\">2,82 \u00b5s (<span class=\"ben1\">1,078\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-7,26%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle_pure_python<\/code><\/td>\n<td class=\"wartn\">217,3 \u00b5s<\/td>\n<td class=\"wartn\">175,2 \u00b5s (<span class=\"ben1s\">1,240\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,38%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pidigits<\/code><\/td>\n<td class=\"wartn\">129,7 ms<\/td>\n<td class=\"wartn\">128,8 ms (<span class=\"ben1\">1,007\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-0,69%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pprint_pformat<\/code><\/td>\n<td class=\"wartn\">1,034 s<\/td>\n<td class=\"wartn\">0,8316 s (<span class=\"ben1s\">1,243\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,55%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pprint_safe_repr<\/code><\/td>\n<td class=\"wartn\">502,5 ms<\/td>\n<td class=\"wartn\">413,3 ms (<span class=\"ben1s\">1,216\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-17,76%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pyflate<\/code><\/td>\n<td class=\"wartn\">323 ms<\/td>\n<td class=\"wartn\">239,9 ms (<span class=\"ben1s\">1,346\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,73%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>python_startup<\/code><\/td>\n<td class=\"wartn\">35,18 ms<\/td>\n<td class=\"wartn\">34,36 ms (<span class=\"ben1\">1,024\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-2,34%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>python_startup_no_site<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">29,09 ms<\/td>\n<td class=\"wartn\">29,23 ms (<span class=\"benvar\">1,005\u00d7 wolniej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">+0,47%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>raytrace<\/code><\/td>\n<td class=\"wartn\">192,8 ms<\/td>\n<td class=\"wartn\">144 ms (<span class=\"ben1s\">1,339\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,34%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_compile<\/code><\/td>\n<td class=\"wartn\">75,55 ms<\/td>\n<td class=\"wartn\">54,3 ms (<span class=\"ben1s\">1,391\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-28,12%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_dna<\/code><\/td>\n<td class=\"wartn\">107,9 ms<\/td>\n<td class=\"wartn\">118,7 ms (<span class=\"ben2s\">1,100\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+10,00%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_effbot<\/code><\/td>\n<td class=\"wartn\">1,588 ms<\/td>\n<td class=\"wartn\">1,94 ms (<span class=\"ben2s\">1,222\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+22,17%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_v8<\/code><\/td>\n<td class=\"wartn\">14,92 ms<\/td>\n<td class=\"wartn\">15,13 ms (<span class=\"ben2\">1,014\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+1,38%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>richards<\/code><\/td>\n<td class=\"wartn\">31,37 ms<\/td>\n<td class=\"wartn\">22,75 ms (<span class=\"ben1s\">1,379\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-27,47%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>richards_super<\/code><\/td>\n<td class=\"wartn\">35,52 ms<\/td>\n<td class=\"wartn\">26,14 ms (<span class=\"ben1s\">1,359\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-26,42%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_fft<\/code><\/td>\n<td class=\"wartn\">217,5 ms<\/td>\n<td class=\"wartn\">163,4 ms (<span class=\"ben1s\">1,331\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-24,89%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_lu<\/code><\/td>\n<td class=\"wartn\">77,23 ms<\/td>\n<td class=\"wartn\">54,56 ms (<span class=\"ben1s\">1,416\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-29,35%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_monte_carlo<\/code><\/td>\n<td class=\"wartn\">48,66 ms<\/td>\n<td class=\"wartn\">36,72 ms (<span class=\"ben1s\">1,325\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-24,54%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_sor<\/code><\/td>\n<td class=\"wartn\">93,32 ms<\/td>\n<td class=\"wartn\">62,58 ms (<span class=\"ben1s\">1,491\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-32,94%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_sparse_mat_mult<\/code><\/td>\n<td class=\"wartn\">3,372 ms<\/td>\n<td class=\"wartn\">2,504 ms (<span class=\"ben1s\">1,347\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,76%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>shortest_path<\/code><\/td>\n<td class=\"wartn\">397,5 ms<\/td>\n<td class=\"wartn\">372,7 ms (<span class=\"ben1\">1,067\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,24%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>spectral_norm<\/code><\/td>\n<td class=\"wartn\">77,96 ms<\/td>\n<td class=\"wartn\">52,58 ms (<span class=\"ben1s\">1,483\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-32,55%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sphinx<\/code><\/td>\n<td class=\"wartn\">617,7 ms<\/td>\n<td class=\"wartn\">552,5 ms (<span class=\"ben1s\">1,118\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,56%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlalchemy_declarative<\/code><\/td>\n<td class=\"wartn\">61,12 ms<\/td>\n<td class=\"wartn\">54,38 ms (<span class=\"ben1s\">1,124\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,02%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlalchemy_imperative<\/code><\/td>\n<td class=\"wartn\">6,798 ms<\/td>\n<td class=\"wartn\">5,959 ms (<span class=\"ben1s\">1,141\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,35%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_normalize<\/code><\/td>\n<td class=\"wartn\">71,32 ms<\/td>\n<td class=\"wartn\">58,48 ms (<span class=\"ben1s\">1,220\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-18,01%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_optimize<\/code><\/td>\n<td class=\"wartn\">33,61 ms<\/td>\n<td class=\"wartn\">27,93 ms (<span class=\"ben1s\">1,203\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,88%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_parse<\/code><\/td>\n<td class=\"wartn\">869,3 \u00b5s<\/td>\n<td class=\"wartn\">621,4 \u00b5s (<span class=\"ben1s\">1,399\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-28,52%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_transpile<\/code><\/td>\n<td class=\"wartn\">1,05 ms<\/td>\n<td class=\"wartn\">0,7834 ms (<span class=\"ben1s\">1,340\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,39%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlite_synth<\/code><\/td>\n<td class=\"wartn\">1,523 \u00b5s<\/td>\n<td class=\"wartn\">1,37 \u00b5s (<span class=\"ben1s\">1,112\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,04%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>subparsers<\/code><\/td>\n<td class=\"wartn\">6,756 ms<\/td>\n<td class=\"wartn\">5,747 ms (<span class=\"ben1s\">1,176\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-14,94%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_expand<\/code><\/td>\n<td class=\"wartn\">258,3 ms<\/td>\n<td class=\"wartn\">220 ms (<span class=\"ben1s\">1,174\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-14,84%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_integrate<\/code><\/td>\n<td class=\"wartn\">11,35 ms<\/td>\n<td class=\"wartn\">9,662 ms (<span class=\"ben1s\">1,175\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-14,90%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_str<\/code><\/td>\n<td class=\"wartn\">149,4 ms<\/td>\n<td class=\"wartn\">124,8 ms (<span class=\"ben1s\">1,198\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,50%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_sum<\/code><\/td>\n<td class=\"wartn\">78,46 ms<\/td>\n<td class=\"wartn\">66,41 ms (<span class=\"ben1s\">1,181\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-15,35%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>telco<\/code><\/td>\n<td class=\"wartn\">4,94 ms<\/td>\n<td class=\"wartn\">4,111 ms (<span class=\"ben1s\">1,202\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,78%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>tomli_loads<\/code><\/td>\n<td class=\"wartn\">1,585 s<\/td>\n<td class=\"wartn\">1,023 s (<span class=\"ben1s\">1,549\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-35,45%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>tornado_http<\/code><\/td>\n<td class=\"wartn\">99,88 ms<\/td>\n<td class=\"wartn\">93,48 ms (<span class=\"ben1\">1,068\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,41%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>typing_runtime_protocols<\/code><\/td>\n<td class=\"wartn\">107 \u00b5s<\/td>\n<td class=\"wartn\">88,07 \u00b5s (<span class=\"ben1s\">1,215\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-17,69%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpack_sequence<\/code><\/td>\n<td class=\"wartn\">48,21 ns<\/td>\n<td class=\"wartn\">31,01 ns (<span class=\"ben1s\">1,555\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-35,68%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpickle<\/code><\/td>\n<td class=\"wartn\">8,886 \u00b5s<\/td>\n<td class=\"wartn\">8,827 \u00b5s (<span class=\"ben1\">1,007\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-0,67%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpickle_list<\/code><\/td>\n<td class=\"wartn\">2,806 \u00b5s<\/td>\n<td class=\"wartn\">2,865 \u00b5s (<span class=\"ben2\">1,021\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+2,11%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpickle_pure_python<\/code><\/td>\n<td class=\"wartn\">157,6 \u00b5s<\/td>\n<td class=\"wartn\">116,6 \u00b5s (<span class=\"ben1s\">1,352\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-26,01%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>urlsafe_base64_small<\/code><\/td>\n<td class=\"wartn\">340,4 \u00b5s<\/td>\n<td class=\"wartn\">206,1 \u00b5s (<span class=\"ben1s\">1,651\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-39,45%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xdsl_constant_fold<\/code><\/td>\n<td class=\"wartn\">24,33 ms<\/td>\n<td class=\"wartn\">20,97 ms (<span class=\"ben1s\">1,160\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,81%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_generate<\/code><\/td>\n<td class=\"wartn\">59,89 ms<\/td>\n<td class=\"wartn\">50,36 ms (<span class=\"ben1s\">1,189\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-15,92%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_iterparse<\/code><\/td>\n<td class=\"wartn\">54,9 ms<\/td>\n<td class=\"wartn\">47,99 ms (<span class=\"ben1s\">1,144\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,59%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_parse<\/code><\/td>\n<td class=\"wartn\">80,08 ms<\/td>\n<td class=\"wartn\">77,91 ms (<span class=\"ben1\">1,028\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-2,71%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_process<\/code><\/td>\n<td class=\"wartn\">42,23 ms<\/td>\n<td class=\"wartn\">34,81 ms (<span class=\"ben1s\">1,213\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-17,58%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><strong>Wynik (\u015brednia geometryczna)<\/strong><\/td>\n<td class=\"wartn\">&#8211;<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,451\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-31,10%<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><span class=\"bench-note\">Jednostki: s \u2014 sekundy; ms \u2014 milisekundy; \u00b5s \u2014 mikrosekundy; ns \u2014 nanosekundy. W ka\u017cdym wierszu oba czasy maj\u0105 t\u0119 sam\u0105 jednostk\u0119. Symbol \u2020 oznacza pomiary wymagaj\u0105ce dodatkowej ostro\u017cno\u015bci.<\/span><\/p>\n<h3>Wydajno\u015b\u0107 wed\u0142ug grup test\u00f3w<\/h3>\n<p>Poni\u017csza tabela podsumowuje testy oznaczone danym znacznikiem w pakiecie pyperformance. Grupy mog\u0105 si\u0119 nak\u0142ada\u0107 i nie obejmuj\u0105 wszystkich test\u00f3w. Grupa <code>math<\/code> zawiera tylko trzy testy; nie jest podsumowaniem wszystkich oblicze\u0144 numerycznych w zestawie.<\/p>\n<div class=\"benchtab-scroll\">\n<table class=\"benchtab\" id=\"pybench-amd-groups\">\n<caption>AMD Ryzen 9 7900 \u2014 komputer stacjonarny: \u015brednia geometryczna wed\u0142ug znacznik\u00f3w<\/caption>\n<thead>\n<tr>\n<th scope=\"col\">Grupa (znacznik)<\/th>\n<th scope=\"col\">Testy dost\u0119pne dla obu wersji<\/th>\n<th scope=\"col\">Python 3.15 wzgl\u0119dem Pythona 3.14<\/th>\n<th scope=\"col\">Odpowiadaj\u0105ca zmiana czasu<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td class=\"nameben\"><code>apps<\/code><\/td>\n<td class=\"wartn\">7<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,130\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,52%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio<\/code><\/td>\n<td class=\"wartn\">21<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,113\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,13%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>math<\/code><\/td>\n<td class=\"wartn\">3<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,238\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,21%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex<\/code><\/td>\n<td class=\"wartn\">4<\/td>\n<td class=\"wartn\"><span class=\"ben1\">1,005\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-0,52%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>serialize<\/code><\/td>\n<td class=\"wartn\">25<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">3,046\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-67,17%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>startup<\/code><\/td>\n<td class=\"wartn\">2<\/td>\n<td class=\"wartn\"><span class=\"ben1\">1,010\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-0,95%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>template<\/code><\/td>\n<td class=\"wartn\">2<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,200\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,66%<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><span class=\"bench-note\">Testy z wynikiem tylko dla Pythona 3.14, wy\u0142\u0105czone z por\u00f3wnania: <code>genshi_text<\/code>, <code>genshi_xml<\/code>. Dostarczone dane nie wyja\u015bniaj\u0105 braku tych wynik\u00f3w dla Pythona 3.15.<\/span><\/p>\n<h2>Mini PC z procesorem Intel Core i3-1315U<\/h2>\n<p>Mini PC Intel jest wyposa\u017cony w procesor Core i3-1315U, pami\u0119\u0107 RAM DDR4 oraz dysk M.2 PCIe Gen4 NVMe. Model i3-1315U jest procesorem mobilnym stosowanym tak\u017ce w laptopach.<\/p>\n<p>Na tym komputerze dost\u0119pne s\u0105 wyniki <strong>118 test\u00f3w dla obu wersji<\/strong>. \u015arednia geometryczna wsp\u00f3\u0142czynnik\u00f3w przyspieszenia Pythona 3.15 wynosi <strong>1,451\u00d7<\/strong>. W znormalizowanym podsumowaniu zestawu odpowiada to skr\u00f3ceniu czasu o oko\u0142o <strong>31,1%<\/strong>. Mediana wsp\u00f3\u0142czynnik\u00f3w przyspieszenia wynosi <strong>1,166\u00d7<\/strong>, co pokazuje, \u017ce niewielka liczba wyj\u0105tkowo du\u017cych przyspiesze\u0144 podnosi wynik og\u00f3lny.<\/p>\n<p>Najwi\u0119ksze przyspieszenia wyst\u0119puj\u0105 w testach <code>base32_large<\/code> (<span class=\"ben1s\">142,255\u00d7 szybciej<\/span>), <code>base85_large<\/code> (<span class=\"ben1s\">62,619\u00d7 szybciej<\/span>), <code>ascii85_large<\/code> (<span class=\"ben1s\">56,825\u00d7 szybciej<\/span>). Warianty tych test\u00f3w kodowania dla ma\u0142ych danych r\u00f3wnie\u017c uzyskuj\u0105 du\u017ce przyspieszenia.<\/p>\n<p>Najwi\u0119ksze wyra\u017ane spowolnienia w u\u015brednionych wynikach wyst\u0119puj\u0105 w testach <code>base16_small<\/code> (<span class=\"ben2s\">1,241\u00d7 wolniej<\/span>), <code>base16_large<\/code> (<span class=\"ben2s\">1,158\u00d7 wolniej<\/span>), <code>regex_effbot<\/code> (<span class=\"ben2s\">1,133\u00d7 wolniej<\/span>). Spadki te s\u0105 szczeg\u00f3lnie istotne dla aplikacji, kt\u00f3re cz\u0119sto wykonuj\u0105 te operacje.<\/p>\n<p>Pomiary Intela wykazywa\u0142y wi\u0119ksz\u0105 zmienno\u015b\u0107. W szczeg\u00f3lno\u015bci wielko\u015b\u0107 przyspieszenia w testach <code>pathlib<\/code>, <code>sphinx<\/code>, <code>scimark_sor<\/code>, <code>scimark_sparse_mat_mult<\/code> i <code>spectral_norm<\/code> zale\u017ca\u0142a od sesji pomiarowej. Niekt\u00f3re pomiary <code>pathlib<\/code> i puli w\u0105tk\u00f3w zawieraj\u0105 wyj\u0105tkowo d\u0142ugie czasy wykonania. U\u015brednienie daje zbiorczy obraz wynik\u00f3w, ale poszczeg\u00f3lne warto\u015bci nie stanowi\u0105 precyzyjnej prognozy dla innego komputera.<\/p>\n<div class=\"benchtab-scroll\">\n<table class=\"benchtab\" id=\"pybench-intel\">\n<caption>Intel Core i3-1315U \u2014 mini PC: u\u015brednione czasy wykonania<\/caption>\n<thead>\n<tr>\n<th scope=\"col\">Test<\/th>\n<th scope=\"col\">Python 3.14.8<\/th>\n<th scope=\"col\">Python 3.15.0<\/th>\n<th scope=\"col\">Zmiana czasu<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td class=\"nameben\"><code>2to3<\/code><\/td>\n<td class=\"wartn\">277,1 ms<\/td>\n<td class=\"wartn\">247,6 ms (<span class=\"ben1s\">1,119\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,65%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>ascii85_large<\/code><\/td>\n<td class=\"wartn\">828,8 ms<\/td>\n<td class=\"wartn\">14,58 ms (<span class=\"ben1s\">56,825\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-98,24%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>ascii85_small<\/code><\/td>\n<td class=\"wartn\">15,57 ms<\/td>\n<td class=\"wartn\">0,4626 ms (<span class=\"ben1s\">33,666\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-97,03%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_generators<\/code><\/td>\n<td class=\"wartn\">280,8 ms<\/td>\n<td class=\"wartn\">230,4 ms (<span class=\"ben1s\">1,219\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-17,96%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_cpu_io_mixed<\/code><\/td>\n<td class=\"wartn\">435,1 ms<\/td>\n<td class=\"wartn\">397,7 ms (<span class=\"ben1\">1,094\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-8,59%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_cpu_io_mixed_tg<\/code><\/td>\n<td class=\"wartn\">423,2 ms<\/td>\n<td class=\"wartn\">411,9 ms (<span class=\"ben1\">1,027\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-2,66%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager<\/code><\/td>\n<td class=\"wartn\">118,4 ms<\/td>\n<td class=\"wartn\">104,9 ms (<span class=\"ben1s\">1,129\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,42%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_cpu_io_mixed<\/code><\/td>\n<td class=\"wartn\">382,7 ms<\/td>\n<td class=\"wartn\">360,1 ms (<span class=\"ben1\">1,063\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-5,90%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_cpu_io_mixed_tg<\/code><\/td>\n<td class=\"wartn\">383,3 ms<\/td>\n<td class=\"wartn\">362 ms (<span class=\"ben1\">1,059\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-5,56%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_io<\/code><\/td>\n<td class=\"wartn\">572,6 ms<\/td>\n<td class=\"wartn\">513,3 ms (<span class=\"ben1s\">1,116\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,36%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_io_tg<\/code><\/td>\n<td class=\"wartn\">561,1 ms<\/td>\n<td class=\"wartn\">509 ms (<span class=\"ben1s\">1,102\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,28%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_memoization<\/code><\/td>\n<td class=\"wartn\">216 ms<\/td>\n<td class=\"wartn\">196,2 ms (<span class=\"ben1s\">1,101\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,16%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_memoization_tg<\/code><\/td>\n<td class=\"wartn\">247,5 ms<\/td>\n<td class=\"wartn\">227,3 ms (<span class=\"ben1\">1,089\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-8,19%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_eager_tg<\/code><\/td>\n<td class=\"wartn\">191,6 ms<\/td>\n<td class=\"wartn\">170,4 ms (<span class=\"ben1s\">1,124\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,07%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_io<\/code><\/td>\n<td class=\"wartn\">593,4 ms<\/td>\n<td class=\"wartn\">517,4 ms (<span class=\"ben1s\">1,147\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,81%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_io_tg<\/code><\/td>\n<td class=\"wartn\">567,9 ms<\/td>\n<td class=\"wartn\">495,3 ms (<span class=\"ben1s\">1,147\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,78%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_memoization<\/code><\/td>\n<td class=\"wartn\">306,3 ms<\/td>\n<td class=\"wartn\">274,5 ms (<span class=\"ben1s\">1,116\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,38%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_memoization_tg<\/code><\/td>\n<td class=\"wartn\">285,9 ms<\/td>\n<td class=\"wartn\">252,9 ms (<span class=\"ben1s\">1,130\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,54%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_none<\/code><\/td>\n<td class=\"wartn\">253,3 ms<\/td>\n<td class=\"wartn\">220,6 ms (<span class=\"ben1s\">1,148\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,92%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>async_tree_none_tg<\/code><\/td>\n<td class=\"wartn\">228,9 ms<\/td>\n<td class=\"wartn\">203,7 ms (<span class=\"ben1s\">1,123\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,98%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio_tcp<\/code><\/td>\n<td class=\"wartn\">582,6 ms<\/td>\n<td class=\"wartn\">539 ms (<span class=\"ben1\">1,081\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-7,48%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio_tcp_ssl<\/code><\/td>\n<td class=\"wartn\">1,537 s<\/td>\n<td class=\"wartn\">1,527 s (<span class=\"ben1\">1,007\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-0,66%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio_websockets<\/code><\/td>\n<td class=\"wartn\">181,7 ms<\/td>\n<td class=\"wartn\">157,8 ms (<span class=\"ben1s\">1,151\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,15%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base16_large<\/code><\/td>\n<td class=\"wartn\">6,161 ms<\/td>\n<td class=\"wartn\">7,132 ms (<span class=\"ben2s\">1,158\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+15,75%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base16_small<\/code><\/td>\n<td class=\"wartn\">298,5 \u00b5s<\/td>\n<td class=\"wartn\">370,3 \u00b5s (<span class=\"ben2s\">1,241\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+24,07%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base32_large<\/code><\/td>\n<td class=\"wartn\">360,9 ms<\/td>\n<td class=\"wartn\">2,537 ms (<span class=\"ben1s\">142,255\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-99,30%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base32_small<\/code><\/td>\n<td class=\"wartn\">7,102 ms<\/td>\n<td class=\"wartn\">0,1753 ms (<span class=\"ben1s\">40,509\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-97,53%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base64_large<\/code><\/td>\n<td class=\"wartn\">8,03 ms<\/td>\n<td class=\"wartn\">1,869 ms (<span class=\"ben1s\">4,297\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-76,73%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base64_small<\/code><\/td>\n<td class=\"wartn\">266,4 \u00b5s<\/td>\n<td class=\"wartn\">191,1 \u00b5s (<span class=\"ben1s\">1,394\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-28,27%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base85_large<\/code><\/td>\n<td class=\"wartn\">298,1 ms<\/td>\n<td class=\"wartn\">4,761 ms (<span class=\"ben1s\">62,619\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-98,40%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>base85_small<\/code><\/td>\n<td class=\"wartn\">5,369 ms<\/td>\n<td class=\"wartn\">0,1747 ms (<span class=\"ben1s\">30,739\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-96,75%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>bench_mp_pool<\/code><\/td>\n<td class=\"wartn\">161,1 ms<\/td>\n<td class=\"wartn\">142,6 ms (<span class=\"ben1s\">1,130\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,47%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>bench_thread_pool<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">1,083 ms<\/td>\n<td class=\"wartn\">1,101 ms (<span class=\"benvar\">1,017\u00d7 wolniej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">+1,71%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>bpe_tokeniser<\/code><\/td>\n<td class=\"wartn\">3,575 s<\/td>\n<td class=\"wartn\">2,782 s (<span class=\"ben1s\">1,285\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-22,19%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>chameleon<\/code><\/td>\n<td class=\"wartn\">11,36 ms<\/td>\n<td class=\"wartn\">9,522 ms (<span class=\"ben1s\">1,193\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,18%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>chaos<\/code><\/td>\n<td class=\"wartn\">47,02 ms<\/td>\n<td class=\"wartn\">35,8 ms (<span class=\"ben1s\">1,314\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,88%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>comprehensions<\/code><\/td>\n<td class=\"wartn\">13,38 \u00b5s<\/td>\n<td class=\"wartn\">9,357 \u00b5s (<span class=\"ben1s\">1,430\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-30,09%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>connected_components<\/code><\/td>\n<td class=\"wartn\">366,4 ms<\/td>\n<td class=\"wartn\">354,6 ms (<span class=\"ben1\">1,033\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-3,22%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>coroutines<\/code><\/td>\n<td class=\"wartn\">16,6 ms<\/td>\n<td class=\"wartn\">13,19 ms (<span class=\"ben1s\">1,258\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-20,53%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>coverage<\/code><\/td>\n<td class=\"wartn\">160,9 ms<\/td>\n<td class=\"wartn\">123,6 ms (<span class=\"ben1s\">1,302\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,19%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>create_gc_cycles<\/code><\/td>\n<td class=\"wartn\">1,711 ms<\/td>\n<td class=\"wartn\">1,428 ms (<span class=\"ben1s\">1,198\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,53%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>crypto_pyaes<\/code><\/td>\n<td class=\"wartn\">57,39 ms<\/td>\n<td class=\"wartn\">45,55 ms (<span class=\"ben1s\">1,260\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-20,62%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deepcopy<\/code><\/td>\n<td class=\"wartn\">206,9 \u00b5s<\/td>\n<td class=\"wartn\">153,5 \u00b5s (<span class=\"ben1s\">1,348\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,81%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deepcopy_memo<\/code><\/td>\n<td class=\"wartn\">20,81 \u00b5s<\/td>\n<td class=\"wartn\">15,72 \u00b5s (<span class=\"ben1s\">1,323\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-24,44%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deepcopy_reduce<\/code><\/td>\n<td class=\"wartn\">2,224 \u00b5s<\/td>\n<td class=\"wartn\">1,71 \u00b5s (<span class=\"ben1s\">1,301\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,14%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>deltablue<\/code><\/td>\n<td class=\"wartn\">2,451 ms<\/td>\n<td class=\"wartn\">1,839 ms (<span class=\"ben1s\">1,333\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-24,99%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>django_template<\/code><\/td>\n<td class=\"wartn\">28,24 ms<\/td>\n<td class=\"wartn\">22,76 ms (<span class=\"ben1s\">1,241\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,42%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>docutils<\/code><\/td>\n<td class=\"wartn\">1,794 s<\/td>\n<td class=\"wartn\">1,661 s (<span class=\"ben1\">1,080\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-7,42%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>fannkuch<\/code><\/td>\n<td class=\"wartn\">309 ms<\/td>\n<td class=\"wartn\">222,6 ms (<span class=\"ben1s\">1,388\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-27,96%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>fastapi_http<\/code><\/td>\n<td class=\"wartn\">351,1 ms<\/td>\n<td class=\"wartn\">320,7 ms (<span class=\"ben1\">1,095\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-8,64%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>float<\/code><\/td>\n<td class=\"wartn\">56,19 ms<\/td>\n<td class=\"wartn\">45,04 ms (<span class=\"ben1s\">1,248\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,85%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>gc_traversal<\/code><\/td>\n<td class=\"wartn\">2,853 ms<\/td>\n<td class=\"wartn\">2,545 ms (<span class=\"ben1s\">1,121\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,81%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>generators<\/code><\/td>\n<td class=\"wartn\">25,72 ms<\/td>\n<td class=\"wartn\">16,94 ms (<span class=\"ben1s\">1,518\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-34,11%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>go<\/code><\/td>\n<td class=\"wartn\">92,33 ms<\/td>\n<td class=\"wartn\">67,09 ms (<span class=\"ben1s\">1,376\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-27,34%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>hexiom<\/code><\/td>\n<td class=\"wartn\">4,839 ms<\/td>\n<td class=\"wartn\">3,443 ms (<span class=\"ben1s\">1,405\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-28,84%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>html5lib<\/code><\/td>\n<td class=\"wartn\">44,02 ms<\/td>\n<td class=\"wartn\">38,9 ms (<span class=\"ben1s\">1,131\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,62%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>json_dumps<\/code><\/td>\n<td class=\"wartn\">7,362 ms<\/td>\n<td class=\"wartn\">5,696 ms (<span class=\"ben1s\">1,292\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-22,63%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>json_loads<\/code><\/td>\n<td class=\"wartn\">17,18 \u00b5s<\/td>\n<td class=\"wartn\">16,01 \u00b5s (<span class=\"ben1\">1,073\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,82%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>k_core<\/code><\/td>\n<td class=\"wartn\">1,977 s<\/td>\n<td class=\"wartn\">1,91 s (<span class=\"ben1\">1,035\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-3,39%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>logging_format<\/code><\/td>\n<td class=\"wartn\">7,111 \u00b5s<\/td>\n<td class=\"wartn\">5,871 \u00b5s (<span class=\"ben1s\">1,211\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-17,44%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>logging_silent<\/code><\/td>\n<td class=\"wartn\">65,46 ns<\/td>\n<td class=\"wartn\">50,39 ns (<span class=\"ben1s\">1,299\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,02%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>logging_simple<\/code><\/td>\n<td class=\"wartn\">6,635 \u00b5s<\/td>\n<td class=\"wartn\">5,423 \u00b5s (<span class=\"ben1s\">1,223\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-18,26%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>mako<\/code><\/td>\n<td class=\"wartn\">7,544 ms<\/td>\n<td class=\"wartn\">6,543 ms (<span class=\"ben1s\">1,153\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,28%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>many_optionals<\/code><\/td>\n<td class=\"wartn\">585,5 \u00b5s<\/td>\n<td class=\"wartn\">513,5 \u00b5s (<span class=\"ben1s\">1,140\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,30%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>mdp<\/code><\/td>\n<td class=\"wartn\">970,9 ms<\/td>\n<td class=\"wartn\">767 ms (<span class=\"ben1s\">1,266\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-21,00%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>meteor_contest<\/code><\/td>\n<td class=\"wartn\">86,82 ms<\/td>\n<td class=\"wartn\">75,98 ms (<span class=\"ben1s\">1,143\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,49%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>nbody<\/code><\/td>\n<td class=\"wartn\">82,94 ms<\/td>\n<td class=\"wartn\">59,03 ms (<span class=\"ben1s\">1,405\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-28,83%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>nqueens<\/code><\/td>\n<td class=\"wartn\">73,7 ms<\/td>\n<td class=\"wartn\">52,9 ms (<span class=\"ben1s\">1,393\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-28,22%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pathlib<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">88,82 ms<\/td>\n<td class=\"wartn\">77,87 ms (<span class=\"benvar\">1,141\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-12,33%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle<\/code><\/td>\n<td class=\"wartn\">9,156 \u00b5s<\/td>\n<td class=\"wartn\">8,568 \u00b5s (<span class=\"ben1\">1,069\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,41%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle_dict<\/code><\/td>\n<td class=\"wartn\">24,76 \u00b5s<\/td>\n<td class=\"wartn\">20,2 \u00b5s (<span class=\"ben1s\">1,226\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-18,43%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle_list<\/code><\/td>\n<td class=\"wartn\">4,1 \u00b5s<\/td>\n<td class=\"wartn\">3,752 \u00b5s (<span class=\"ben1\">1,093\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-8,48%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pickle_pure_python<\/code><\/td>\n<td class=\"wartn\">246,7 \u00b5s<\/td>\n<td class=\"wartn\">206 \u00b5s (<span class=\"ben1s\">1,198\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,50%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pidigits<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">163,1 ms<\/td>\n<td class=\"wartn\">163,7 ms (<span class=\"benvar\">1,003\u00d7 wolniej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">+0,34%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pprint_pformat<\/code><\/td>\n<td class=\"wartn\">1,186 s<\/td>\n<td class=\"wartn\">0,9441 s (<span class=\"ben1s\">1,257\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-20,42%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pprint_safe_repr<\/code><\/td>\n<td class=\"wartn\">585,2 ms<\/td>\n<td class=\"wartn\">476,1 ms (<span class=\"ben1s\">1,229\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-18,64%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>pyflate<\/code><\/td>\n<td class=\"wartn\">343,9 ms<\/td>\n<td class=\"wartn\">267 ms (<span class=\"ben1s\">1,288\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-22,36%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>python_startup<\/code><\/td>\n<td class=\"wartn\">31,09 ms<\/td>\n<td class=\"wartn\">30,31 ms (<span class=\"ben1\">1,026\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-2,53%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>python_startup_no_site<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">23,24 ms<\/td>\n<td class=\"wartn\">23,16 ms (<span class=\"benvar\">1,003\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-0,32%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>raytrace<\/code><\/td>\n<td class=\"wartn\">223,5 ms<\/td>\n<td class=\"wartn\">164,2 ms (<span class=\"ben1s\">1,361\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-26,54%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_compile<\/code><\/td>\n<td class=\"wartn\">94,27 ms<\/td>\n<td class=\"wartn\">76,29 ms (<span class=\"ben1s\">1,236\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,07%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_dna<\/code><\/td>\n<td class=\"wartn\">136,6 ms<\/td>\n<td class=\"wartn\">135,9 ms (<span class=\"ben1\">1,005\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-0,49%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_effbot<\/code><\/td>\n<td class=\"wartn\">1,71 ms<\/td>\n<td class=\"wartn\">1,937 ms (<span class=\"ben2s\">1,133\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+13,30%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex_v8<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">16,61 ms<\/td>\n<td class=\"wartn\">15,72 ms (<span class=\"benvar\">1,057\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-5,36%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>richards<\/code><\/td>\n<td class=\"wartn\">32,36 ms<\/td>\n<td class=\"wartn\">24,82 ms (<span class=\"ben1s\">1,304\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,29%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>richards_super<\/code><\/td>\n<td class=\"wartn\">36,74 ms<\/td>\n<td class=\"wartn\">28,57 ms (<span class=\"ben1s\">1,286\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-22,22%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_fft<\/code><\/td>\n<td class=\"wartn\">212,5 ms<\/td>\n<td class=\"wartn\">158,1 ms (<span class=\"ben1s\">1,344\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-25,58%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_lu<\/code><\/td>\n<td class=\"wartn\">69,29 ms<\/td>\n<td class=\"wartn\">52,02 ms (<span class=\"ben1s\">1,332\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-24,92%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_monte_carlo<\/code><\/td>\n<td class=\"wartn\">50,14 ms<\/td>\n<td class=\"wartn\">36,87 ms (<span class=\"ben1s\">1,360\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-26,47%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_sor<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">103,8 ms<\/td>\n<td class=\"wartn\">57,98 ms (<span class=\"benvar\">1,790\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-44,12%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>scimark_sparse_mat_mult<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">3,352 ms<\/td>\n<td class=\"wartn\">2,368 ms (<span class=\"benvar\">1,415\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-29,35%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>shortest_path<\/code><\/td>\n<td class=\"wartn\">403,9 ms<\/td>\n<td class=\"wartn\">382 ms (<span class=\"ben1\">1,057\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-5,40%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>spectral_norm<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">77,37 ms<\/td>\n<td class=\"wartn\">50,15 ms (<span class=\"benvar\">1,543\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-35,18%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sphinx<\/code> <sup>\u2020<\/sup><\/td>\n<td class=\"wartn\">812,6 ms<\/td>\n<td class=\"wartn\">725,8 ms (<span class=\"benvar\">1,120\u00d7 szybciej<\/span> <sup>\u2020<\/sup>)<\/td>\n<td class=\"wartn\"><span class=\"benvar\">-10,69%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_normalize<\/code><\/td>\n<td class=\"wartn\">85,23 ms<\/td>\n<td class=\"wartn\">68,7 ms (<span class=\"ben1s\">1,241\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-19,39%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_optimize<\/code><\/td>\n<td class=\"wartn\">40,98 ms<\/td>\n<td class=\"wartn\">34,54 ms (<span class=\"ben1s\">1,186\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-15,71%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_parse<\/code><\/td>\n<td class=\"wartn\">949,5 \u00b5s<\/td>\n<td class=\"wartn\">737,1 \u00b5s (<span class=\"ben1s\">1,288\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-22,37%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlglot_v2_transpile<\/code><\/td>\n<td class=\"wartn\">1,18 ms<\/td>\n<td class=\"wartn\">0,9215 ms (<span class=\"ben1s\">1,281\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-21,92%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sqlite_synth<\/code><\/td>\n<td class=\"wartn\">1,832 \u00b5s<\/td>\n<td class=\"wartn\">1,704 \u00b5s (<span class=\"ben1\">1,075\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-6,96%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>subparsers<\/code><\/td>\n<td class=\"wartn\">7,632 ms<\/td>\n<td class=\"wartn\">6,672 ms (<span class=\"ben1s\">1,144\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,57%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_expand<\/code><\/td>\n<td class=\"wartn\">337,3 ms<\/td>\n<td class=\"wartn\">290,4 ms (<span class=\"ben1s\">1,162\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,93%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_integrate<\/code><\/td>\n<td class=\"wartn\">14,24 ms<\/td>\n<td class=\"wartn\">12,71 ms (<span class=\"ben1s\">1,120\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,72%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_str<\/code><\/td>\n<td class=\"wartn\">196,8 ms<\/td>\n<td class=\"wartn\">171,8 ms (<span class=\"ben1s\">1,145\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,67%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>sympy_sum<\/code><\/td>\n<td class=\"wartn\">100,9 ms<\/td>\n<td class=\"wartn\">88,9 ms (<span class=\"ben1s\">1,135\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,88%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>telco<\/code><\/td>\n<td class=\"wartn\">5,451 ms<\/td>\n<td class=\"wartn\">4,658 ms (<span class=\"ben1s\">1,170\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-14,55%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>tomli_loads<\/code><\/td>\n<td class=\"wartn\">1,679 s<\/td>\n<td class=\"wartn\">1,096 s (<span class=\"ben1s\">1,532\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-34,74%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>tornado_http<\/code><\/td>\n<td class=\"wartn\">105,5 ms<\/td>\n<td class=\"wartn\">102,7 ms (<span class=\"ben1\">1,027\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-2,61%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>typing_runtime_protocols<\/code><\/td>\n<td class=\"wartn\">127,7 \u00b5s<\/td>\n<td class=\"wartn\">110,8 \u00b5s (<span class=\"ben1s\">1,153\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,24%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpack_sequence<\/code><\/td>\n<td class=\"wartn\">37,45 ns<\/td>\n<td class=\"wartn\">31,33 ns (<span class=\"ben1s\">1,195\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,32%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpickle<\/code><\/td>\n<td class=\"wartn\">9,764 \u00b5s<\/td>\n<td class=\"wartn\">9,974 \u00b5s (<span class=\"ben2\">1,022\u00d7 wolniej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben2\">+2,15%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpickle_list<\/code><\/td>\n<td class=\"wartn\">3,154 \u00b5s<\/td>\n<td class=\"wartn\">3,144 \u00b5s (<span class=\"ben1\">1,003\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-0,31%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>unpickle_pure_python<\/code><\/td>\n<td class=\"wartn\">161,4 \u00b5s<\/td>\n<td class=\"wartn\">123,3 \u00b5s (<span class=\"ben1s\">1,310\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-23,64%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>urlsafe_base64_small<\/code><\/td>\n<td class=\"wartn\">417,2 \u00b5s<\/td>\n<td class=\"wartn\">235,3 \u00b5s (<span class=\"ben1s\">1,773\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-43,58%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xdsl_constant_fold<\/code><\/td>\n<td class=\"wartn\">35,23 ms<\/td>\n<td class=\"wartn\">31,16 ms (<span class=\"ben1s\">1,131\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-11,56%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_generate<\/code><\/td>\n<td class=\"wartn\">65,1 ms<\/td>\n<td class=\"wartn\">56,4 ms (<span class=\"ben1s\">1,154\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-13,37%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_iterparse<\/code><\/td>\n<td class=\"wartn\">69,49 ms<\/td>\n<td class=\"wartn\">61,14 ms (<span class=\"ben1s\">1,137\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-12,02%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_parse<\/code><\/td>\n<td class=\"wartn\">108,3 ms<\/td>\n<td class=\"wartn\">103,5 ms (<span class=\"ben1\">1,047\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-4,44%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>xml_etree_process<\/code><\/td>\n<td class=\"wartn\">46,24 ms<\/td>\n<td class=\"wartn\">38,3 ms (<span class=\"ben1s\">1,207\u00d7 szybciej<\/span>)<\/td>\n<td class=\"wartn\"><span class=\"ben1\">-17,17%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><strong>Wynik (\u015brednia geometryczna)<\/strong><\/td>\n<td class=\"wartn\">&#8211;<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,451\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-31,10%<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><span class=\"bench-note\">Jednostki: s \u2014 sekundy; ms \u2014 milisekundy; \u00b5s \u2014 mikrosekundy; ns \u2014 nanosekundy. W ka\u017cdym wierszu oba czasy maj\u0105 t\u0119 sam\u0105 jednostk\u0119. Symbol \u2020 oznacza pomiary wymagaj\u0105ce dodatkowej ostro\u017cno\u015bci.<\/span><\/p>\n<h3>Wydajno\u015b\u0107 wed\u0142ug grup test\u00f3w<\/h3>\n<p>Poni\u017csza tabela podsumowuje testy oznaczone danym znacznikiem w pakiecie pyperformance. Grupy mog\u0105 si\u0119 nak\u0142ada\u0107 i nie obejmuj\u0105 wszystkich test\u00f3w. Grupa <code>math<\/code> zawiera tylko trzy testy; nie jest podsumowaniem wszystkich oblicze\u0144 numerycznych w zestawie.<\/p>\n<div class=\"benchtab-scroll\">\n<table class=\"benchtab\" id=\"pybench-intel-groups\">\n<caption>Intel Core i3-1315U \u2014 mini PC: \u015brednia geometryczna wed\u0142ug znacznik\u00f3w<\/caption>\n<thead>\n<tr>\n<th scope=\"col\">Grupa (znacznik)<\/th>\n<th scope=\"col\">Testy dost\u0119pne dla obu wersji<\/th>\n<th scope=\"col\">Python 3.15 wzgl\u0119dem Pythona 3.14<\/th>\n<th scope=\"col\">Odpowiadaj\u0105ca zmiana czasu<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td class=\"nameben\"><code>apps<\/code><\/td>\n<td class=\"wartn\">7<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,108\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-9,77%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>asyncio<\/code><\/td>\n<td class=\"wartn\">21<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,114\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-10,27%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>math<\/code><\/td>\n<td class=\"wartn\">3<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,204\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,97%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>regex<\/code><\/td>\n<td class=\"wartn\">4<\/td>\n<td class=\"wartn\"><span class=\"ben1\">1,037\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-3,60%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>serialize<\/code><\/td>\n<td class=\"wartn\">25<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">3,044\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-67,15%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>startup<\/code><\/td>\n<td class=\"wartn\">2<\/td>\n<td class=\"wartn\"><span class=\"ben1\">1,015\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-1,43%<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\"><code>template<\/code><\/td>\n<td class=\"wartn\">2<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,196\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1\">-16,40%<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><span class=\"bench-note\">Testy z wynikiem tylko dla Pythona 3.14, wy\u0142\u0105czone z por\u00f3wnania: <code>dask<\/code>, <code>genshi_text<\/code>, <code>genshi_xml<\/code>, <code>sqlalchemy_declarative<\/code>, <code>sqlalchemy_imperative<\/code>. Dostarczone dane nie wyja\u015bniaj\u0105 braku tych wynik\u00f3w dla Pythona 3.15.<\/span><\/p>\n<h2>Przyspieszenie operacji kodowania<\/h2>\n<p>Najwi\u0119ksze zmiany tworz\u0105 wyra\u017any wzorzec. U\u015bredniony wynik <code>base32_large<\/code> wskazuje przyspieszenie oko\u0142o <strong>135\u00d7 na AMD<\/strong> i <strong>142\u00d7 na Intelu<\/strong>. Testy Ascii85 i Base85 dla du\u017cych danych uzyskuj\u0105 przyspieszenia oko\u0142o <strong>57\u201369\u00d7<\/strong>, a odpowiadaj\u0105ce im testy dla ma\u0142ych danych \u2014 oko\u0142o <strong>31\u201341\u00d7<\/strong>. S\u0105 to wyspecjalizowane operacje, kt\u00f3rych przyspieszenie znacznie przewy\u017csza wyniki wi\u0119kszo\u015bci pozosta\u0142ych test\u00f3w.<\/p>\n<p>Wzorzec ten jest zgodny z informacjami zawartymi w <a href=\"https:\/\/docs.python.org\/3.15\/whatsnew\/3.15.html#base64-binascii\" rel=\"noopener\" target=\"_blank\">opisie zmian w Pythonie 3.15<\/a>: implementacje Base32 oraz Ascii85\/Base85 przepisano w j\u0119zyku C, a Base64 r\u00f3wnie\u017c zoptymalizowano. Zmiany te stanowi\u0105 prawdopodobne wyja\u015bnienie obserwowanych przyspiesze\u0144, chocia\u017c pomiary nie pozwalaj\u0105 oddzieli\u0107 wp\u0142ywu poszczeg\u00f3lnych zmian w kodzie. Testy mierz\u0105 czas wykonania, wi\u0119c nie potwierdzaj\u0105 oszcz\u0119dno\u015bci pami\u0119ci.<\/p>\n<div class=\"benchtab-scroll\">\n<table class=\"benchtab\" id=\"pybench-sensitivity\">\n<caption>Wp\u0142yw wyj\u0105tkowo du\u017cych przyspiesze\u0144 kodowania na wynik og\u00f3lny<\/caption>\n<thead>\n<tr>\n<th scope=\"col\">Wyb\u00f3r test\u00f3w<\/th>\n<th scope=\"col\">Przyspieszenie na AMD<\/th>\n<th scope=\"col\">Przyspieszenie na Intelu<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td class=\"nameben\">Wszystkie testy dost\u0119pne dla obu wersji<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,451\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,451\u00d7 szybciej<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\">Bez sze\u015bciu test\u00f3w Ascii85 \/ Base32 \/ Base85<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,203\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,197\u00d7 szybciej<\/span><\/td>\n<\/tr>\n<tr>\n<td class=\"nameben\">Mediana przyspieszenia poszczeg\u00f3lnych test\u00f3w<\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,184\u00d7 szybciej<\/span><\/td>\n<td class=\"wartn\"><span class=\"ben1s\">1,166\u00d7 szybciej<\/span><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>Po wy\u0142\u0105czeniu tych sze\u015bciu test\u00f3w pozostaje <strong>116 por\u00f3wnywalnych test\u00f3w na AMD<\/strong> i <strong>112 na Intelu<\/strong>. \u015arednie geometryczne wsp\u00f3\u0142czynnik\u00f3w przyspieszenia wynosz\u0105 odpowiednio <strong>1,203\u00d7<\/strong> i <strong>1,197\u00d7<\/strong>. W takim znormalizowanym podsumowaniu odpowiada to skr\u00f3ceniu czasu o oko\u0142o <strong>17%<\/strong>. Jest to przydatny dodatkowy punkt odniesienia dla aplikacji, kt\u00f3re rzadko korzystaj\u0105 z Base32 lub Base85.<\/p>\n<h2>Znaczenie wynik\u00f3w dla wydajno\u015bci aplikacji<\/h2>\n<p>Grupa test\u00f3w ze znacznikiem <code>apps<\/code> uzyskuje przyspieszenie oko\u0142o <strong>1,13\u00d7 na AMD<\/strong> i <strong>1,11\u00d7 na Intelu<\/strong>. Renderowanie szablon\u00f3w przyspiesza oko\u0142o <strong>1,20\u00d7 na obu komputerach<\/strong>. Wyniki wskazuj\u0105 na zauwa\u017calne korzy\u015bci w niekt\u00f3rych zadaniach wykonywanych g\u0142\u00f3wnie w Pythonie. Jednocze\u015bnie pokazuj\u0105, dlaczego \u015bredniego wyniku 1,45\u00d7 nie mo\u017cna bezpo\u015brednio przenosi\u0107 na ka\u017cd\u0105 aplikacj\u0119.<\/p>\n<p>Wyniki kodowania s\u0105 zr\u00f3\u017cnicowane: du\u017cym przyspieszeniom Base32 i Base85 towarzysz\u0105 wolniejsze operacje Base16. Podobnie grupa wyra\u017ce\u0144 regularnych uzyskuje jedynie niewielkie przyspieszenie og\u00f3lne, podczas gdy <code>regex_effbot<\/code> dzia\u0142a wolniej na obu komputerach, a <code>regex_dna<\/code> \u2014 na AMD. \u015arednie czasy uruchamiania pozostaj\u0105 bliskie wynikowi referencyjnemu. Praktyczny efekt zale\u017cy od operacji wykonywanych przez aplikacj\u0119.<\/p>\n<p><a href=\"https:\/\/docs.python.org\/3.15\/whatsnew\/3.15.html#build-changes\" rel=\"noopener\" target=\"_blank\">Oficjalna dokumentacja Pythona<\/a> opisuje tak\u017ce interpreter wykorzystuj\u0105cy wywo\u0142ania ogonowe, stosowany w oficjalnych 64-bitowych wydaniach dla Windows. W metadanych badanych wersji zapisano r\u00f3\u017cne identyfikatory kompilatora. Por\u00f3wnanie dotyczy kompletnych kompilacji interpretera, dlatego poszczeg\u00f3lnych przyspiesze\u0144 nie mo\u017cna przypisa\u0107 wy\u0142\u0105cznie JIT, kompilatorowi ani konkretnej zmianie w interpreterze.<\/p>\n<p>Oba komputery uzyskuj\u0105 niemal identyczny wynik dla ca\u0142ego zestawu, ale r\u00f3\u017cni\u0105 si\u0119 sprz\u0119tem i liczb\u0105 test\u00f3w dost\u0119pnych dla obu wersji. Nie oznacza to wi\u0119kszych korzy\u015bci dla kt\u00f3rego\u015b producenta procesor\u00f3w.<\/p>\n<h2>Uwagi dotycz\u0105ce pomiar\u00f3w<\/h2>\n<p>W ka\u017cdej sesji pomiary danego testu s\u0105 wykonywane przez 20 proces\u00f3w roboczych. Dla wi\u0119kszo\u015bci test\u00f3w zachowano 60 warto\u015bci czasu, a dla ka\u017cdego testu uruchamiania \u2014 200. Przy obliczaniu zbiorczych czas\u00f3w \u015brednie z poszczeg\u00f3lnych sesji maj\u0105 tak\u0105 sam\u0105 wag\u0119. Warto\u015bci odstaj\u0105ce pozostawiono w danych, a testy bez wyniku dla jednej z wersji pomini\u0119to zamiast przypisywa\u0107 im wynik szacunkowy.<\/p>\n<p>Symbol \u2020 jest stosowany, gdy wsp\u00f3\u0142czynnik por\u00f3wnania zmienia si\u0119 mi\u0119dzy sesjami o wi\u0119cej ni\u017c 10%, \u015brednie przyspieszenie zmienia si\u0119 w spowolnienie lub odwrotnie albo wsp\u00f3\u0142czynnik zmienno\u015bci zestawu pomiar\u00f3w przekracza 25%. Wsp\u00f3\u0142czynnik zmienno\u015bci to stosunek odchylenia standardowego do \u015bredniej, wyra\u017cony w procentach. Oznaczenie to przyj\u0119to na potrzeby artyku\u0142u; nie jest ono formalnym testem istotno\u015bci statystycznej. Wskazuje \u015brednie, kt\u00f3rych wielko\u015b\u0107, a przy ma\u0142ych r\u00f3\u017cnicach r\u00f3wnie\u017c kierunek zmiany, jest mniej wiarygodna.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Por\u00f3wnano wydajno\u015b\u0107 Python 3.14.8 i Python 3.15.0 na komputerach z procesorem serii AMD Ryzen 7000 oraz procesorem Intel Core 13. generacji. Jeden z nich u\u017cywany jest w komputerach stacjonarnych, drugi \u2013 w laptopach lub komputerach mini PC. W obu przypadkach u\u017cyto 64-bitowych kompilacji Pythona dzia\u0142aj\u0105cych w systemie Windows 11. Testy przeprowadzono za pomoc\u0105 pakietu pyperformance [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":1118,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[12],"tags":[184,138,185,139,137,140],"class_list":["post-1117","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-artykuly","tag-amd","tag-amd-ryzen-serii-7000","tag-intel","tag-intel-core-13-generacji","tag-python","tag-test-wydajnosci"],"_links":{"self":[{"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/posts\/1117","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/comments?post=1117"}],"version-history":[{"count":0,"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/posts\/1117\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/media\/1118"}],"wp:attachment":[{"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/media?parent=1117"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/categories?post=1117"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lewoniewski.info\/blog\/wp-json\/wp\/v2\/tags?post=1117"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}