Kana hungwaru hwekugadzira huchionekwa serwendo kubva kuA kuenda kuB, cloud computing sevhisi nhandare yendege kana chiteshi chechitima chinomhanya, uye edge komputa itekisi kana bhasikoro rakagovaniswa. Edge komputa iri padyo nedivi revanhu, zvinhu, kana data masosi. Inotora chikuva chakavhurika chinosanganisa kuchengetedza, computation, network kuwana, uye application musimboti kugona kupa masevhisi kune vashandisi vari pedyo. Kuenzaniswa neyakaiswa pakati pegore computing masevhisi, edge komputa inogadzirisa matambudziko akadai seyakareba latency uye yakanyanya convergence traffic, ichipa rutsigiro rwuri nani kune chaiyo-nguva uye bandwidth-inoda masevhisi.
Moto weChatGPT wakamisa mutsva wekuvandudzwa kweAI, kukurumidza kunyura kweAI munzvimbo dzakawanda dzekushanda dzakadai semaindasitiri, zvitoro, dzimba dzakangwara, maguta akachenjera, etc. Nhamba yakawanda yedata inoda kuchengetwa uye kuverengwa pa kuguma kwekushandisa, uye kuvimba negore chete hakuchakwanisi kuzadzisa icho chaidiwa, komputa komputa inovandudza kiromita yekupedzisira yeAI application. Pasi pemutemo wenyika wekusimbaradza hupfumi hwedhijitari, China gore computing yapinda munguva yekusimudzira, kudiwa kwemakomputa kwawedzera, uye kubatanidzwa kwemupendero wegore uye kupera kwave nzira yakakosha yekushanduka-shanduka mune ramangwana.
Edge komputa musika kukura 36.1% CAGR mumakore mashanu anotevera
Indasitiri yemakomputa yemupendero yapinda padanho rekusimudzira, sezvinoratidzwa nekusiyana zvishoma nezvishoma kwevanopa masevhisi, kukura kwemusika kukura, uye kumwe kuwedzera kwenzvimbo dzekushandisa. Panyaya yehukuru hwemusika, dhata kubva kuIDC yekutevera rondedzero inoratidza kuti musika wese weiyo edge computing maseva muChina yakasvika US $ 3.31 bhiriyoni muna 2021, uye saizi yemusika yese yeedge computing maseva muChina inotarisirwa kukura pakukura kwepagore. mwero we22.2% kubva 2020 kusvika 2025. Sullivan anofanotaura kukura kwemusika komputa komputa muChina inotarisirwa kusvika RMB 250.9 bhiriyoni muna 2027, ine CAGR ye36.1% kubva 2023 kusvika 2027.
Edge komputa eco-indasitiri inobudirira
Edge komputa parizvino iri padanho rekutanga rekubuda, uye miganhu yebhizinesi mumaindasitiri ketani haina kujeka. Kune vatengesi vega, zvinofanirwa kufunga nezve kubatanidzwa nemamiriro ebhizinesi, uye zvakare zvinodiwa kuve nekugona kuchinjika kune shanduko yemamiriro ebhizinesi kubva padanho rehunyanzvi, uye zvakare zvinodiwa kuve nechokwadi chekuti kune dhigirii repamusoro. kuenderana nemidziyo yehardware, pamwe nekugona kweinjiniya kuita mapurojekiti.
Iyo yekumucheto komputa indasitiri cheni yakakamurwa kuva chip vatengesi, algorithm vatengesi, Hardware mudziyo vagadziri, uye mhinduro vanopa. Vatengesi veChip vanowanzogadzira arithmetic machipisi kubva kumucheto-kudivi kuenda kumucheto-kurutivi kuenda kune gore-parutivi, uye nekuwedzera kune epamucheto-machipisi, ivo zvakare vanogadzira makadhi ekumhanyisa uye kutsigira software yekuvandudza mapuratifomu. Algorithm vatengesi vanotora komputa vision algorithms semusimboti wekuvaka general kana customized algorithms, uye kune zvakare mabhizinesi anovaka algorithm mall kana kudzidzisa uye kusunda mapuratifomu. Vatengesi vezvishandiso vari kushingaira kuisa mari mune edge computing zvigadzirwa, uye chimiro che edge computing zvigadzirwa zvinogara zvichipfumiswa, zvishoma nezvishoma zvichigadzira yakazara stack ye edge komputa zvigadzirwa kubva ku chip kuenda kumuchina wese. Vagadziri vemhinduro vanopa software kana software-hardware-yakabatanidzwa mhinduro kune chaiwo maindasitiri.
Edge komputa indasitiri yekushandisa inomhanyisa
Mumunda weguta rakangwara
Kuongorora kwakadzama kwezvivakwa zvemudhorobha pari zvino kunowanzo shandiswa munzira yekuongorora nemaoko, uye iyo yemanyorero yekuongorora maitiro ine matambudziko ekudhura-nguva-yakanyanya uye kudhura kwevashandi, kutsamira pamaitiro emunhu, kuvharika kwakashata uye kuwanda kwekuongorora, uye hurombo husina kunaka. control. Panguva imwecheteyo maitiro ekuongorora akarekodha huwandu hukuru hwe data, asi izvi zviwanikwa zve data hazvina kushandurwa kuita data assets yekugonesa bhizinesi. Nekushandisa tekinoroji yeAI kune nharembozha yekuongorora mamiriro, bhizinesi rakagadzira hutongi hwemaguta AI yakangwara yekuongorora mota, iyo inotora matekinoroji akadai seInternet Yezvinhu, cloud computing, AI algorithms, uye inotakura midziyo yehunyanzvi yakadai semakamera akakwirira-tsanangudzo, pa- bhodhi inoratidza, uye AI padivi maseva, uye inosanganisa nzira yekuongorora ye "intelligent system + intelligent machine + staff assistance". Inosimudzira shanduko yehutongi hwemadhorobha kubva kuvashandi-yakanyanya kuenda kuhungwaru hwemuchina, kubva pakutonga kwesimba kuenda kuongororo yedata, uye kubva pakupindura kusingaite kuenda kukutsvaga kwakasimba.
Mumunda wenzvimbo yekuvaka yakangwara
Edge computing-yakavakirwa akangwara ekuvaka saiti mhinduro dzinoshandisa kusanganisa kwakadzama kweAI tekinoroji kune yechinyakare indasitiri yekuvaka kuchengetedza basa rekutarisa, nekuisa mupendero AI yekuongorora terminal panzvimbo yekuvaka, kupedzisa yakazvimirira tsvagiridzo uye kusimudzira yekuona AI algorithms yakavakirwa pahungwaru vhidhiyo. analytics tekinoroji, nguva yakazara yekuona zviitiko zvinofanirwa kuonekwa (semuenzaniso, kuona kana kusapfeka ngowani), kupa vashandi, nharaunda, chengetedzo uye kumwe kuchengetedzwa kwenzvimbo yengozi uye masevhisi echiyeuchidzo chearamu, uye kutora danho reKuzivikanwa kwevasina kuchengetedzeka. zvinhu, AI yakangwara kuchengetedza, kuchengetedza manpower mutengo, kusangana nevashandi uye kuchengetedza kuchengetedza zvivakwa zvinodiwa zvenzvimbo dzekuvaka.
Mumunda wekutakura kwakangwara
Cloud-side-end architecture yave iyo yekutanga paradigm yekutumirwa kwezvikumbiro muindasitiri yekufambisa ine hungwaru, nedivi regore rinotarisira manejimendi epakati uye chikamu chekugadzirisa data, rutivi rwemupendero rwunonyanya kupa kumucheto-kurutivi data kuongororwa uye sarudzo yekombuta. -kugadzira kugadzirisa, uye rutivi rwekupedzisira rinonyanya kuita basa rekuunganidza data rebhizimisi.
Mune chaiwo mamiriro akadai semota-mugwagwa kurongeka, holographic mharadzano, kutyaira otomatiki, uye njanji traffic, kune huwandu hukuru hwezvishandiso zvakasiyana-siyana zvinowanikwa, uye zvishandiso izvi zvinoda manejimendi ekuwana, kubuda kwekutonga, alarm kugadzirisa, uye kushanda uye kugadzirisa kugadzirisa. Edge komputa inogona kupatsanura uye kukunda, kushandura hombe kuita diki, kupa muchinjika-layer protocol shanduko mabasa, kuwana yakabatana uye yakagadzikana kuwana, uye kunyange yekudyidzana kutonga kweheterogeneous data.
Mumunda wekugadzira maindasitiri
Kugadzira Maitiro Ekugadzirisa Mamiriro: Parizvino, huwandu hukuru hwemadhisiki ekugadzira masisitimu anoganhurirwa nekusakwana kwedata, uye kugona kwese kwemidziyo uye kumwe kuverengera kwedata data kwakapusa, zvichiita kuti zviome kushandisa kuita zvakanaka. Edge computing chikuva chakavakirwa pamidziyo yeruzivo modhi kuti uwane semantic nhanho yekugadzira sisitimu yakatwasuka kutaurirana uye yakatwasuka kutaurirana, zvichibva pane chaiyo-nguva data kuyerera kwekugadzirisa nzira yekuunganidza uye kuongorora nhamba yakakura yemunda chaiyo-nguva data, kuwana modhi-yakavakirwa mutsara wekugadzira. Multi-data source information fusion, kupa yakasimba data rutsigiro rwekuita sarudzo mune discrete yekugadzira system.
Equipment Predictive Maintenance Scenario: Maintenance yezvigadzirwa zvemaindasitiri zvakakamurwa kuva marudzi matatu: reparative kugadzirisa, kudzivirira kuchengetedza, uye kufanotaura kugadzirisa. Restorative gadziriso ndeye ex post facto gadziriso, yekudzivirira kuchengetedza, uye fungidziro yekuchengetedza ndeye ex-ante kuchengetedza, yekutanga kunoenderana nenguva, mashandiro emidziyo, mamiriro esaiti, uye zvimwe zvinhu zvekugara uchigadzirisa michina, zvakanyanya kana zvishoma zvichibva pamunhu. ruzivo, iyo yekupedzisira kuburikidza nekuunganidza data sensor, chaiyo-nguva yekutarisa mamiriro ekushanda emidziyo, zvichienderana neindasitiri yemhando yekuongorora data, uye kunyatsofanotaura kana kukundikana kunoitika.
Mamiriro ekuongorora mhando yeindasitiri: nzvimbo yekuongorora maindasitiri ndiyo yekutanga otomatiki optical inspection (AOI) fomu mundima yekuongorora mhando, asi kuvandudzwa kweAOI kusvika parizvino, mukuonekwa kwakakanganiswa uye zvimwe zvakaomarara, nekuda kwekukanganisa kwakasiyana. yemhando, kudhirowa kwechimiro hakuna kukwana, adaptive algorithms yakashata kuwedzera, mutsara wekugadzira unovandudzwa kazhinji, kutama kwealgorithm hakusi kuchinjika, uye zvimwe zvinhu, iyo yechinyakare AOI hurongwa hwave hwakaoma kusangana nekuvandudzwa kwezvinodiwa mutsetse wekugadzira. Naizvozvo, iyo AI indasitiri yemhando yekuongorora algorithm papuratifomu inomiririrwa nekudzidza kwakadzama + diki sampuli yekudzidza zvishoma nezvishoma iri kutsiva yechinyakare yekuona yekuongorora chirongwa, uye iyo AI indasitiri yemhando yekuongorora chikuva yapfuura nematanho maviri echinyakare muchina kudzidza algorithms uye yakadzika yekudzidza algorithms.
Nguva yekutumira: Oct-08-2023