I-Edge Computing Kwiimoto Ezizimeleyo

I-Edge computing iguqula izithuthi ezizimeleyo ngokuvumela ukucutshungulwa kwedatha enkulu ngexesha langempela ukusuka kwiisensa ezifana neLiDAR, i-radar, kunye neekhamera. Ngokungafaniyo ne-cloud computing yendabuko, exhomekeke kwiiseva ezikude, i-edge computing icwangcisa idatha kwindawo yezithuthi okanye kwizixhobo ezikufutshane, inciphisa i-latency kwaye iphucula ukuthembeka. Oku kubalulekile kwiimoto eziziqhuba ngokwazo, apho izigqibo zemizuzwana embalwa zinokuthintela ukungqubana kwaye ziqinisekise ukhuseleko lwabakhweli. Ngokudibanisa unxibelelwano lwe-5G, i-AI, kunye ne-vehicle-to-everything (V2X), i-edge computing inika amandla izithuthi ukuba zisebenzisane ngokungenamthungo nezinye iimoto, iziseko zophuhliso, kunye nabahambi ngeenyawo, idala inkqubo yothutho ekrelekrele edibeneyo.

Kutheni oku kubalulekile? Ukuqhuba ngokuzimeleyo kufuna uhlalutyo olukhawulezileyo lwedatha yesenzi ukuze kuhambeke kwiindawo ezintsonkothileyo, ukusuka kwizitalato zasezidolophini ezixakekileyo ukuya kwiindlela zasemaphandleni ezinonxibelelwano olunamabala. Ukucubungula i-low-latency emphethweni kunciphisa ukuxhomekeka kwiiseva zamafu, kunciphisa iindleko ze-bandwidth kunye nokuphucula ukusebenza kwiindawo ezinonxibelelwano olubuthathaka lwenethiwekhi. Umzekelo, iinkqubo ze-V2X zivumela izithuthi ukuba zabelane ngohlaziyo lwethrafikhi ngexesha langempela, ziphucula ulawulo lwethrafikhi kunye nokusebenza kakuhle kwepetroli. Ukongeza, i-edge AI ixhasa ukufunda komatshini ngaphakathi, ivumela izithuthi ukuba zilungelelanise neemeko eziguquguqukayo ngaphandle konxibelelwano oluqhubekayo lwamafu.

Olu kudibana kwetekhnoloji kuvula indlela yezixeko ezikrelekrele kunye neendlela ezikhuselekileyo. Ngokusebenzisa ubukrelekrele bomda, izithuthi ezixhunyiweyo zinokusebenzisana neeyunithi ezisecaleni kwendlela (ii-RSU) kunye neziseko ezikrelekrele, kunciphisa ukuxinana kunye nokukhutshwa komoya. Ukuhlanganiswa kweenethiwekhi ze-5G kuqinisekisa uhlalutyo lwexesha langempela, ngelixa amanyathelo okhuseleko lwe-cyber afana ne-blockchain ekhusela idatha eyimfihlo. Njengoko izithuthi ezizimeleyo ziguqukela ekuzimeleni ngokupheleleyo, i-edge computing imi njengelitye lembombo, iqhuba ubuchule kwii-apps zokuhamba, ukuqokelelana kweelori, kunye nokugcinwa kwangaphambili. Eli candelo lihlola indlela ezi nkqubela phambili eziguqula ngayo uthutho, zibeka inqanaba lekamva apho ukuhamba okukrelekrele kusebenza kakuhle kwaye kukhuselekile.

I-Edge Computing iziphucula njani iimoto ezizimeleyo?

I-Edge computing iguqula izithuthi ezizimeleyo ngokucubungula idatha yasekuhlaleni, ngokungafaniyo ne-cloud computing, exhomekeke kwiiseva ezikude. Olu tshintsho lubalulekile kwiimoto eziziqhuba ngokwazo, apho izigqibo ze-low latency zibalulekile kukhuseleko kunye nokusebenza. Ngokuphatha idatha ye-sensor evela kwi-LiDAR, i-radar, kunye neekhamera ezikwi-board okanye kwizixhobo ezikufutshane, izithuthi zifezekisa ukucutshungulwa ngexesha langempela, kunciphisa ukulibaziseka okunokubangela iingozi. Umzekelo, ukufumanisa umntu ohamba ngeenyawo kufuna uhlalutyo olukhawulezileyo, oluya kubonelela nge-edge computing ngaphandle kokulibaziseka kokudluliselwa kwelifu.

Iingenelo zicacile:

Ukubambezeleka okuphantsi kakhulu: Ivumela ukuhambahamba kwemizuzwana embalwa kunye nokuthintela ukungqubana.

Ukusebenza kakuhle kwe-bandwidth: Kunciphisa idatha ethunyelwa kwilifu, kunciphisa iindleko ze-bandwidth.

Ukuthembeka: Kuqinisekisa ukusebenza kwiindawo ezinonxibelelwano olubi lwenethiwekhi, njengeendlela zasemaphandleni.

I-Edge computing ixhasa usetyenziso oluphambili ekuqhubeni ngokuzimela. Inika amandla unxibelelwano phakathi kwezithuthi (V2X), ivumela izithuthi ukuba zabelane ngedatha nezinye iimoto (V2V), iziseko zophuhliso (V2I), okanye abahamba ngeenyawo (V2P) ngeenethiwekhi ze-5G. Oku kuphucula ulawulo lwezithuthi, kunciphisa ukuxinana kunye nokuphucula ukusebenza kakuhle kwepetroli. Umzekelo, i-edge intelligence isebenza ngedatha yokudibanisa i-sensor ukuze imephu yeendawo ezingqongileyo ngexesha langempela, ivumela ukuhamba ngokuchanekileyo. Ukongeza, i-edge AI isebenzisa iimodeli zokufunda koomatshini ngaphakathi, ixhasa ukubonwa kwezithintelo kunye nokuqhuba okuguqukayo ngaphandle kokuxhomekeka rhoqo kwilifu.

Xa kuthelekiswa ne-cloud computing, i-edge computing inika amaxesha okuphendula akhawulezayo kunye nokuxhomekeka okuncitshisiweyo kwi-centralized computing, okubaluleke kakhulu kwizithuthi ezixhunyiweyo kwiindawo eziguquguqukayo. Ikwaxhasa izixeko ezikrelekrele ngokudibanisa neeyunithi zendlela (ii-RSU) zohlaziyo lwethrafikhi ngexesha langempela. Ngokuphucula uhlalutyo lwedatha kwindawo, i-edge computing inciphisa ukubambezeleka kwenethiwekhi, okwenza kube yintsika yeenkqubo zothutho ezikrelekrele. Le ndlela iqinisekisa ukuba izithuthi ezizimeleyo zisebenza ngokufanelekileyo, ngokukhuselekileyo, nangokungabizi kakhulu, ivula indlela yee-apps zokuhamba ezinokwandiswa kunye neziseko ezikrelekrele ezomeleleyo.

Ukunika amandla iiTekhnoloji ze-Edge Computing kwiimoto ezizimeleyo

I-Edge computing kwizithuthi ezizimeleyo ixhomekeke kubuchwepheshe obuphambili ukubonelela ngenkqubo yexesha langempela kunye nonxibelelwano olungenamthungo. Iinethiwekhi ze-5G, ezidibene ne-multi-access edge computing (MEC), zibalulekile, zibonelela ngonxibelelwano oluphantsi lwe-latency kwiinkqubo ze-vehicle-to-everything (V2X). I-5G ivumela ukusebenzisana kwe-V2V, V2I, kunye ne-V2P, ivumela izithuthi ukuba zabelane ngedatha yethrafikhi neeyunithi zendlela (ii-RSU) okanye iziseko ezikrelekrele, okuphucula ulawulo lwethrafikhi kunye nokuthintela ukungqubana. I-MEC yabucala imisela inkxaso kubaqhubi beenqwelo-moya, ngelixa i-MEC kawonkewonke iqinisekisa ukuhlanganiswa okubanzi kwesixeko esikrelekrele, okwenza izithuthi eziqhagamshelweyo ziphendule ngakumbi.

Ubukrelekrele bokwenziwa (AI) kunye nokufunda koomatshini kuqhuba ubukrelekrele bomda, ukucubungula idatha yesenzi esivela kwiLiDAR, iradar, kunye neekhamera ngexesha langempela. Iimodeli zokufunda nzulu, ezisebenza kwizixhobo ezikhethekileyo ezifana neNVIDIA DRIVE okanye iQualcomm Snapdragon, zivumela uhlalutyo lwexesha langempela lokuhamba kunye nokuchonga izithintelo. Ezi qonga ziphucula izixhobo zokubala, zilungelelanisa ukusebenza kunye nokusetyenziswa kwamandla. Umzekelo, i-edge AI ihlalutya iimeko zendlela ngoko nangoko, inciphisa ukuxhomekeka kwi-cloud computing.

I-sensor fusion idibanisa idatha evela kwimithombo emininzi ukuze kwenziwe imaphu echanekileyo yokusingqongileyo. Ngokudibanisa i-LiDAR, i-radar, kunye neekhamera, izithuthi zifezekisa ukwenza izigqibo ngokuchanekileyo ngexesha langempela. I-Edge computing ihluza le datha kwindawo, inciphisa ukusetyenziswa kwe-bandwidth kwaye inciphise umthwalo kwiiseva zamafu. Oku kubalulekile kwiinkqubo zothutho ezikrelekrele (ITS), apho uhlalutyo lwedatha olusebenzayo luxhasa ukuhamba okukrelekrele.

Ubuchwepheshe obuphambili buquka:

I-5G kunye ne-MEC: Vumela unxibelelwano lwe-V2X olukhawulezayo noluthembekileyo.

Umphetho we-AI: Inika amandla ukufunda komatshini okusemoyeni ukuze kufumaneke iimpendulo ezitshintshatshintshayo.

Ukuhlanganiswa kweSensor: Idibanisa idatha ukuze kubekho ukuzulazula okuchanekileyo, ngexesha langempela.

Iingenelo ze-Edge Computing kwiiNkqubo zeZithuthi eziZimeleyo

I-Edge computing iguqula izithuthi ezizimeleyo ngokuvumela ukucutshungulwa ngexesha langempela, inika iingenelo ezibalulekileyo zokhuseleko, ukusebenza kakuhle, kunye nokukhula. Ngokucutshungulwa kwedatha yesenzwa evela kwiLiDAR, i-radar, kunye neekhamera zasekuhlaleni, i-edge computing iqinisekisa iimpendulo ze-low-latency ezibalulekileyo ekuthinteleni ukungqubana. Umzekelo, ukubona imiqobo okanye abahambi ngeenyawo kwi-milliseconds kuphucula ukhuseleko lwezithuthi, kunciphisa iingozi kwiindawo eziguquguqukayo njengezitalato zasezidolophini. Ulawulo lwezithuthi luyaphucuka njengoko ubukrelekrele be-edge buxhasa unxibelelwano lwezithuthi-ukuya-konke-konke (V2X). Nge-V2V kunye ne-V2I, izithuthi zabelana ngedatha yexesha langempela neeyunithi zendlela (ii-RSU) kunye neziseko ezikrelekrele, ziphucula ukuhamba kwezithuthi kwaye zinciphise ukuxinana. Oku kukhokelela ekusebenzeni kakuhle kwepetroli, ukunciphisa iindleko kunye nokukhutshwa kwegesi kwizithuthi ezixhunyiweyo. I-Edge computing ikwanciphisa ukusetyenziswa kwe-bandwidth ngokucoca idatha yasekuhlaleni, ukunciphisa ukuxhomekeka kwi-cloud computing kunye nokunciphisa iindleko ze-bandwidth. Kwizixeko ezikrelekrele, i-edge computing ivumela iinkqubo zothutho ezikrelekrele ezikhula ngokukhawuleza (i-ITS). Ixhasa ii-apps zokuhamba kunye nokuqokelelana kwelori, apho izithuthi zidibana khona ukuze zonge ipetroli kwaye zilungelelanise uthutho. Ngokucutshungulwa kwedatha e-edge, iinkqubo zokuqhuba ezizimeleyo zihlala zithembekile nakwiindawo ezinonxibelelwano olubi lwenethiwekhi, ukuqinisekisa ukusebenza okuzinzileyo. Iingenelo eziphambili ziquka: Ukhuseleko Oluphuculweyo: Uhlalutyo lwexesha langempela lokufumanisa ingozi ngoko nangoko. Iinzuzo Zokusebenza Kakuhle: Ukuphucula ithrafikhi kunye nokusebenza kakuhle kwepetroli nge-V2X. Ukwanda: Ixhasa ukuhlanganiswa kwesixeko esikrelekrele kunye nolawulo lweenqwelo-mafutha.



Izicelo Ezisebenzayo kunye neMizekelo Yoshishino
I-Edge computing inika amandla usetyenziso oluguqulayo kwi izithuthi ezizimelayo, evumela ukucutshungulwa ngexesha langempela kwiindlela ezahlukeneyo zokusetyenziswa. Kwintengiso iinqwelo-moya ezizimeleyoiinkampani ezifana neWaymo harness ubukrelekrele bomda ngokusebenzisa okuphambili iPC efakiweyo iinkqubo zokucutshungulwa idatha yesenzi ukusuka I-LiDAR, iradakunye iikhamera, ukwenza kube lula ukuhanjiswa kwempahla kunye nokuhanjiswa kwayo. Ezi zithuthi zixhomekeke umphetho we-AI ngenxa uhlalutyo lwexesha langempela, ukuqinisekisa ukuhambahamba okusebenzayo kunye ukuthintela ukungqubana kwiindawo ezitshintshatshintshayo. Umzekelo, iPC encinci engenafeni ivumela iiveni zokuhambisa ukuba zilungelelanise iindlela ngoko nangoko, zinqumle ukusetyenziswa kwepetroli kunye namaxesha okuhanjiswa.


Ngaphakathi izixeko ezikrelekrele, ikhompyutha esemaphethelweni kuphucula ulawulo lwezithuthi ngokusebenzisa isithuthi-kwinto yonke (V2X) iinkqubo. Iiyunithi ezisecaleni kwendlela (ii-RSU) kwaye iziseko zophuhliso ezikrelekrele, ixhaswa yi-robust ikhompyutha ye-rackmount yemizi-mveliso izisombululo, ukucubungula idatha kwindawo yakho ukuze ukwazi V2I unxibelelwano, ukulungelelanisa imiqondiso yendlela kunye nokunciphisa ukuxinana. I-Apollo Go yeBaidu iyadibanisa 5G kwaye i-rackmount PC iinkqubo zokuxhasa ukuqhuba ngokuzimela kwiindawo zasezidolophini, ukunyusa ukhuseleko lwezithuthi kunye nokuhamba kwezithuthi. Olu manyano lukhuthaza ukwandiswa iinkqubo zothutho ezikrelekrele (ITS), inkxaso ii-apps zokuhamba uhlaziyo lwethrafikhi ngexesha langempela.


I-Intel's Mobileye ibonisa ikhompyutha esemaphethelweni kwizithuthi zabathengi, ukusebenzisa ukufunda nzulu kwi Ikhompyuter encinci yeShishini ngenxa ukufunyanwa kwemiqobo kunye nokuhambahamba. Ngokucwangciswa ukuhlanganiswa kwenzwa idatha ekwibhodi, iMobileye iyanciphisa ukubambezeleka kwenethiwekhi kwaye ukuxhomekeka kwilifu, ukuqinisekisa ukuthembeka kwiindawo ezihlwempuzekileyo uqhagamshelo lwenethiwekhiLe ndlela ikwanciphisa iindleko zebhendi, ukwenza izithuthi eziqhagamshelweyo kufikeleleka ngakumbi kwaye kusebenza kakuhle.




Izicelo eziphambili ziquka:


Iinqanawa Ezizimeleyo: Ukulungiswa kwendlela ngexesha langempela lokuhambisa izinto.
Ukuhlanganiswa kweSmart City: I-V2X yokulungelelanisa izithuthi kunye nokonga ipetroli.
Izithuthi zabathengi: I-AI esecaleni kwendlela yokukhangela ekhuselekileyo.


Le mizekelo igxininisa indlela i-edge computing enika ngayo amandla okusebenza ngobuchule, kunye neenkampani ezifana ne-NVIDIA kunye ne-Tesla eziqhubela phambili usetyenziso olusekelwe kumgca. Ngokudibanisa iinethiwekhi ze-5G, ukhuseleko lwe-cyber, kunye nohlalutyo lwedatha, i-edge computing iqinisekisa ukuba izithuthi ezizimeleyo zisebenza kakuhle, zikhuselekile, kwaye zilungele ukusetyenziswa ngokubanzi ezidolophini.




Iqonga Lekhompyutha Eliqinileyo Linika Amandla Okuqhuba Ngokuzimeleyo




Njengoko ukuqhuba ngokuzimela kufikeleleka ngokupheleleyo kwiindawo eziphambili zasezidolophini zaseHangzhou, ukuqhubela phambili kwezixhobo ezikrelekrele kuye kwathatha indawo ephambili kwinkqubela phambili yezobuchwepheshe. Ukuze sihlangabezane nale mfuno, sazisa iqonga le-AI elinobunkunkqele elilungele ukuqhuba ngokuzimela kunye neenkqubo zothutho ezikrelekrele, ezixhotyiswe nge-CPU kunye ne-GPU enamandla.



Amandla eKhompyutha aPhezulu


Iqonga le-SIN-3412-R680E liqhutywa zii-CPU ze-Intel ze-12th/13th Gen, ezixhasa ukuya kuthi ga kwi-cores ezingama-24 kunye nemisonto engama-32 yokuphatha imisebenzi enzima ngokufanelekileyo. Ineendawo ezimbini zememori ze-64GB ECC, eziqinisekisa inkxaso ethembekileyo kwimisebenzi ebalulekileyo. Kwindawo yokugcina, inikezela nge-M.2 2280 M key slot exhasa ii-Gen4 x4 NVMe SSDs, kunye neendlela ezimbini zokugcina ezishushu ezinokuthi zitshintshwe kwiphaneli yangaphambili: enye ye-2.5" SATA HDD/SSD kunye netreyi ye-M.2 2280 Gen4 x4 NVMe enokukhetha kuyo, ejongana neemfuno ezahlukeneyo zokugcina.



Ujongano Olubanzi kunye Nezakhono Zokwandisa


I-SIN-3412-R680E ibonelela ngeseti etyebileyo yee-interfaces ze-I/O ezikwibhodi, kuquka ii-ports ezintandathu ze-USB 3.2 Gen2, ii-ports ezintathu ze-2.5G Ethernet/Gigabit Ethernet, kunye ne-port ye-10G Ethernet ekhethiweyo. Ikwabandakanya ii-interfaces ezimbini ze-RS-232/422/485 ezinokucwangciswa kwisoftware ukuze ziqhagamshelwe ngeendlela ezahlukeneyo. Ukwandisa, iqonga line-slot enye ye-PCIe x16 kunye nobubanzi be-slot obuyi-6.5mm obubekelwe amakhadi e-GPU. I-bracket yokutshixa ye-GPU entsha iqinisekisa uzinzo, ngelixa ifreyimu eyahlukileyo efunxa ukutshaqeka inciphisa ukungcangcazela okuphezulu, iqinisekisa ukusebenza okuthembekileyo kuzo zombini iinkqubo ezisendleleni nangaphandle kwendlela. Ixhasa ii-GPU ze-NVIDIA® RTX A4000, A5000, A6000, 6000 Ada, kunye nee-RTX 40 series ezikhethiweyo.




Ezinye iindlela zokwandisa


Eli qonga libonelela ngeeslots ezintathu ezongezelelweyo ze-PCIe zamakhadi okubamba iikhamera ze-GMSL2/industrial okanye amanye amakhadi okwandisa, iislots ezimbini ze-mini PCIe ezinde zokwandisa izibuko ze-CAN bus/COM, kunye neslots enye ye-M.2 B yeemodyuli ze-4G LTE, 5G NR, okanye ze-WiFi, ezibonelela ngokwenziwa ngokwezifiso okuguquguqukayo.




Uyilo oluTsha lweThermal


Uyilo lokupholisa olungenafeni oluqinisekisiweyo luqinisekisa ukusebenza okuzinzileyo kweCPU kunye neGPU kumaqondo obushushu ukusuka kwi -25°C ukuya kwi-60°C. Oku kunciphisa ingxolo, kuphucula ukuthembeka, kwaye kwandisa ubomi, okwenza kube yeyona ifanelekileyo yokuqhuba ngokuzimela kwiindawo zasezidolophini.



Amathuba okufaka isicelo


Iqonga le-SIN-3412-R680E, elinokusebenza okugqwesileyo kunye nokwandiswa okuguquguqukayo, linika inkxaso eqinileyo yokuqhuba ngokuzimela kunye nokuthutha ngobuchule. Lisebenza ngokuthembekileyo kwiindawo ezintsonkothileyo, lihlangabezana neemfuno zokuhamba ngobuchule ezidolophini. Njengoko izixeko ezikrelekrele zitshintsha, eli qonga liyimbangela ephambili yetekhnoloji yezithuthi ezizimeleyo, linikezela ngoyilo olutsha kunye nokusebenza okuthembekileyo kwikamva lokuthutha ngobuchule. Ukuba unemibuzo okanye ufuna ukwazi okungakumbi, zive ukhululekile ukunxibelelana nathi!