Buga na Buga na 59 na martabar kwamfutoci 500 tare da mafi girman aikin na duniya. A cikin sabon bugu, manyan goma sun haɗa da sabbin gungu uku, waɗanda suka tabbatar da matsayi na farko, na uku da na goma.
Wuri na farko ya tafi zuwa sabon Frontier, wanda yake a Ma'aikatar Makamashi ta Amurka ta Oak Ridge Laboratory National. gungu yana da kusan nau'in sarrafawa miliyan 9 (AMD EPYC 64C 2GHz CPU, AMD Instinct MI250X Accelerator) kuma yana ba da aikin petaflops 1102, wanda shine kusan sau uku jagoran baya. A lokaci guda, ƙarfin wutar lantarki na Frontier ya ragu da kashi 30% fiye da shugaban da ya gabata.
Tarin Fugaku na Jafananci da aka gina tare da masu sarrafa ARM (158976 nodes dangane da Fujitsu A64FX SoC, sanye take da 8.2-core 2,2GHz Armv48-A SVE CPU) ya koma matsayi na biyu. An gudanar da gungu na Fugaku a Cibiyar RIKEN don Binciken Jiki da Kimiyya kuma yana ba da aikin 442 petaflops. Tsarin aiki shine Red Hat Enterprise Linux.
Sabuwar LUMI cluster ta samu matsayi na uku. An shirya shi a Cibiyar Supercomputing ta Turai (EuroHPC) a Finland kuma yana ba da wasan kwaikwayo na petaflops 151. Tarin ya dogara ne akan dandamalin HPE Cray EX235a iri ɗaya a matsayin sabon jagora a cikin martaba, amma ya haɗa da na'urori masu sarrafawa miliyan 1,1 (AMD EPYC 64C 2GHz, AMD Instinct MI250X accelerator, Slingshot-11 network). Hakanan ana amfani da wannan dandali a cikin gungun Adastra na Faransa, wanda ke matsayi na 10 kuma yana da nau'ikan sarrafawa guda 319.000 (jimlar ayyukan 46 petaflops).
Gungu taron, IBM ya tura shi a Oak Ridge National Laboratory (Amurka), ya fadi zuwa matsayi na hudu. Tarin yana gudanar da Red Hat Enterprise Linux, ya haɗa da nau'ikan sarrafawa miliyan 2,4 (ta amfani da 9-core IBM Power22 3,07C 22 GHz CPUs da NVIDIA Tesla V100 accelerators), yana ba da 148 petaflops na aiki.
Amma ga supercomputers na gida, ƙungiyoyi Chervonenkis, Galushkin da Lyapunov wanda Yandex ya kirkira sun fadi daga wurare na 19, 36 da 40. zuwa wurare 22, 40 da 43. An tsara waɗannan gungu don magance matsalolin koyon inji da ba da aikin 21,5, 16, da 12,8 petaflops, bi da bi. Gungu suna gudanar da Ubuntu 16.04 kuma an sanye su da na'urori masu sarrafawa na AMD EPYC 7xxx da NVIDIA A100 GPUs: gungu na Chervonenkis yana da nodes 199 (193K AMD EPYC 7702 64C 2GH cores da 1592 NVIDIA A100 80G GPUs 136GH da 134 NVDIA A7702 64G GPUs, 2GH 1088K100GH80K137K 130des da 7662 NVIDIA A64 2G GPUs), Lyapunov - 1096 nodes (100 dubu AMD EPYC 40 XNUMXC XNUMXGHz cores da XNUMX NVIDIA AXNUMX XNUMXG GPUs).
Ƙungiyar Christofari Neo da Sberbank ya tura ta faɗi daga matsayi na 43 zuwa 46. Christofari Neo yana gudanar da NVIDIA DGX OS 5 (Ubuntu Edition) kuma yana ba da 11,9 petaflops na aiki. Tarin yana da fiye da 98k cores bisa AMD EPYC 7742 64C 2.25GHz CPU kuma ya zo tare da NVIDIA A100 80GB GPU. Rukunin na biyu na Sberbank (Christofari) ya tashi daga matsayi na 72 zuwa 80 a cikin rabin shekara.
Abubuwan da suka fi ban sha'awa sune masu zuwa. Rarraba ta adadin manyan kwamfutoci a ƙasashe daban-daban:
- China: 173 (173 - rabin shekara da suka wuce). Gabaɗaya, gungu na kasar Sin yana samar da kashi 12% na duk abin da ake samarwa (watanni shida da suka gabata, 17,5%)
- Amurka: 127 (149). An ƙiyasta jimlar ayyuka a kashi 47,3% na jimlar aikin ƙima (watanni shida da suka gabata - 32,5%)
- Japan: 34 (32). Jimlar yawan aiki - 14,3%
- Jamus: 31 (26). Gabaɗaya Yawan aiki - 11,5%
- Faransa: 22 (19)
- Kanada 14 (11)
- Burtaniya: 12 (11)
- Rasha 7 (7)
- Netherlands: 6 (11)
- Italiya: 6 (6)
- Brazil 6 (5)
- Koriya ta Kudu 6 (7)
- Saudi Arabia 6 (6)
- Poland 5 (4)
- Ostiraliya 5 (3)
- Sweden 5 (4)
- Switzerland 4 (3)
- Finland: 4 (3).
A cikin martabar tsarin aiki da aka yi amfani da su a cikin manyan kwamfutoci, Linux kawai ya riƙe nasa tsawon shekaru biyar:
- 47,8% (51,6%) ba su dalla-dalla rarrabawar
- 18.2% (18%) amfani da CentOS
8,8% (7,6%) - RHEL - 8% (7%) CrayLinux
- 5,2% (5,4%) - Ubuntu
- 3,8% (4%) - SUSE
- 0.8% (0%) - Alma Linux
- 0,6% (0%) - RockyLinux
- 0,2% (0,2%) - Linux na kimiyya
Matsakaicin iyakar aiki don shigar da Top500 bai canza ba har tsawon watanni 6 kuma yana da 1,65 petaflops (watanni shida da suka gabata, gungun 56 sun nuna aikin 1,65 petaflops, yanzu - 20). Shekaru uku da suka wuce gungu 272 ne kawai suka nuna aiki akan petaflops, shekaru hudu da suka gabata 138, shekaru biyar da suka gabata 94). Don Top100, ƙofar shiga ya karu daga 4,78 zuwa 5,39 petaflops;
yawan amfanin ƙasa gabaɗaya na duk tsarin a cikin ranking ya karu daga 3,04 zuwa 4,40 exaflops a cikin watanni 6 (shekaru biyu da suka gabata shine 1,650 exaflops kuma shekaru biyar da suka gabata shine 566 petaflops). Tsarin da ya rufe matsayi na yanzu yana matsayi na 464 a cikin fitowar ƙarshe.
Ana sa ran za a fitar da sabon sigar madadin ƙimar Graph 500 na tsarin tari nan gaba kaɗan. Green500, HPCG (High-Performance Conjugate Gradient) da HPL-AI ana haɗe su tare da Top500 kuma suna nunawa a cikin babban matsayi na Top500.
A ƙarshe, idan kuna sha'awar ƙarin sani game da shi, zaku iya tuntuɓar cikakkun bayanai A cikin mahaɗin mai zuwa.