I-NVIDIA Jetson AGX Thor: Imininingwane, Ikhithi, kanye Namanani

I-Jetson AGX Thor ifika namamojula e-T5000 ne-T4000 kanye nekhithi yokuthuthukisa egxile kumarobhothi athuthukile.
Kufika ku-2.070 TFLOPS FP4, 7,5x ukwanda kwe-AI kanye nokuthuthukiswa okungu-3,5x kokusebenza kahle nge-Orin.
Ukuxhumana okubanzi: PCIe Gen5, USB-C/USB-A, HDMI/DP, 5GbE kanye ne-QSFP28 4x25GbE kukhithi ye-dev.
Gcwalisa i-ecosystem nge-NVIDIA Isaac, Metropolis, neHoloscan, futhi yamukelwe abaholi abafana ne-Amazon Robotics ne-Boston Dynamics.

Ipulatifomu ye-NVIDIA Jetson AGX Thor yamarobhothi ne-AI

I-NVIDIA yethula ikhithi yonjiniyela ye-Jetson AGX Thor kanye namamojula ayo okukhiqiza, a I-AI edge computing platform yakhelwe ukukhiqiza amarobhothi, Logistics, ezokuthutha, ezempilo, ezolimo kanye nokudayisa. Inhloso icacile: Ukuletha i-AI ekhiqizayo kanye nokucabanga kwe-multimodal kumarobhothi ukuze bakwazi ukuxhumana nabantu kanye nezindawo ngesikhathi sangempela.

Abaphathi be-NVIDIA bayakugcizelela lokho Labo abakha izinhlelo zamarobhothi bathembele kumapulatifomu abo ukuze bahlanganise ukusebenza okuphezulu nokusebenza kahle kwamandla. futhi ngaleyo ndlela sebenzisa amamodeli e-AI amaningi emaphethelweni

Amandla nokusebenza kahle kwe-AI ebonakalayo emaphethelweni

I-Jetson AGX Thor Module AI Ukusebenza

Ukweqa uma kuqhathaniswa nesizukulwane sangaphambilini kuyamangaza: kufika izikhathi ezingu-7,5 ngaphezulu kwamandla ekhompuyutha e-AI kanye nokuthuthukiswa kwe 3,5 izikhathi ukonga amandla ngaphezulu ngokumelene neJetson AGX Orin. Inkundla isilungele ukusebenza kwe-multi-AI, kufakiwe umbono-ulimi-isenzo amamodeli batch yakamuva njenge Isaac GR00T N1.5 namamodeli olimi amahle.

Imodyuli Jetson T5000 ihlanganisa i-GPU NVIDIA Blackwell con 2.560 CUDA Cores kanye 96 Tensor Cores, CPU ye 14 ARM cores at 2,6 GHz, futhi I-128GB LPDDR5X phezulu I-273 GB / s kwe-bandwidth. Ngezibalo, ifinyelela 2.070 TFLOPS ku-FP4 ngokusetshenziswa kwereferensi ye 130 W, isisekelo esiqinile samarobhothi adinga umbono osheshayo kanye nokwenza izinqumo.

Okuqukethwe okukhethekile - Chofoza Lapha Ungayivula kanjani ikhibhodi yeLenovo Legion 5?

Kulabo ababeka phambili ukusebenza kahle, i Jetson T4000 inikeza up I-1.200 TFLOPS ngeBlackwell GPU evela 1.525 CUDA Cores kanye 64 Tensor Cores, CPU ye 12 ARM cores at 2,6 GHz, futhi I-64GB LPDDR5X ukugcina i 273 GB/s umkhawulokudonsa, enephrofayili yamandla nxazonke 70 W.

Idizayinelwe i-AI yomzimba, womabili amamojula Zikuvumela ukuthi ucubungule kanyekanye okokufaka okuvela kumakhamera, i-LiDAR, nezinye izinzwa, usebenzise amamodeli akhiqizayo nokulawula, futhi uphendule ngokubambezeleka okuphansi., okuthile okubalulekile kuma-humanoid, ama-drones, izingalo zezimboni noma abasizi bokuhlinzwa.

Ikhithi Yokuthuthukisa i-Jetson AGX Thor kanye Nokuxhumana Kwezinga Eliphezulu

Jetson AGX Thor Development Kit kanye Motherboard

El I-Jetson AGX Thor Developer Kit ihlanganisa imojuli T5000 kanye nebhodi lomama elinokuxhumana okuningi ukuze kusheshiswe ukuhlola, ukuhlanganiswa kwezinzwa, nokuphakelwa kokukhiqiza. Iphakheji ihlanganisa ukuchithwa okusebenzayo nokudla ukuze ukuthunyelwa kusheshe.

Isitoreji/Ukunweba: izikhala ezimbili M.2 (PCIe Gen5 x4 kanye ne-Gen5 x1) yama-NVMe SSD noma amanye ama-peripheral.
Amaports: 2x I-USB-A 3.1ewu, 2x I-USB-C 3.1, I-HDMI i-2.0b y BonisaPort 1.4a.
RediRj45 I-Ethernet 5 GbE y QSFP28 kuze kufike ku-4×25 GbE.
Ukuxhumana okwengeziwe: M.2 ye-Wi-Fi, izihloko zokuphina, umsindo kanye isikhombimsebenzisi sekhamera ilungele ukwenza i-prototyping.

Okuqukethwe okukhethekile - Chofoza Lapha Ungalifaka kanjani ikhadi lezithombe?

Konke lokhu kuhlanganiswe ku-a I-compact form factor engu-243,19 × 112,40 × 56,88 mm kanye nobubanzi be-TDP obungu-40 kuya ku-130 W, afanele amabhentshi okuhlola, amarobhothi ahambayo kanye neziteshi zokuqinisekisa zelabhorethri.

I-Ecosystem, amamodeli kanye nokutholwa embonini

I-NVIDIA Isaac, Metropolis, kanye ne-Holoscan Ecosystem

UJetson Thor uthembela kusitaki sesoftware UNVIDIA Jetson ye-AI engokomzimba neyerobhothi, ehambisana nobukhulu Izinhlaka ze-AI ezikhiqizayo kanye nokuhlanganiswa komdabu nge NVIDIA Isaac (ukulingisa kanye namarobhothi), Metropolis (umbono wekhompyutha) kanye IHoloscan (ukucubungula inzwa yesikhathi sangempela).

I-Jetson ecosystem ikhule phakathi neminyaka ukuze ihlanganise ndawonye ngaphezu kwezigidi ezingu-2 zonjiniyela kanye nendwangu ye Abalingani abangu-150+ ku-hardware, isofthiwe nezinzwa. Nge Jetson Orin kakade isebenza ngaphezu 7.000 amaklayenti Kuyo yonke imikhakha eminingi; U-Thor uthatha isinyathelo esilandelayo ukuze anike amandla izinhlelo eziyinkimbinkimbi, kuhlanganise humanoids kanye amarobhothi okuhlinza.

Izinkampani ezifana I-Agility Robotics, i-Amazon Robotics, i-Boston Dynamics, i-Caterpillar, i-Figure, i-Hexagon, i-Medtronic ne-Meta vele usebenze nomsamo, nabanye bathanda UJohn Deere kanye ne-OpenAI hlola ukufaneleka kwabo kumaphrojekthi we-AI aphathekayo. Umbono owabiwe yilokho amamodeli athuthukile asebenza erobhothini ngokwalo, ukunciphisa ukubambezeleka nokuncika kwamafu.

Ngokombono womkhiqizo kugqanyisiwe lokho sebenzisa amamodeli akhiqizayo esikalini emaphethelweni ivumela amarobhothi ukuthi abone, acabange futhi enze ngokuzimela okukhulu, nokuthi Inhlanganisela yamandla ekhompyutha nokusebenza kahle kuyisihluthulelo sokusebenza ngokuphepha ezindaweni eziguquguqukayo zokuthutha kanye nezimboni.

Okuqukethwe okukhethekile - Chofoza Lapha Ukuthuthukiswa kwe-PC: Igayidi yobuchwepheshe yokuthuthukisa ikhompuyutha yakho

Ukutholakala kanye namanani

El I-Jetson AGX Thor Developer Kit Manje ingathengwa kusuka Amadola ka-3.499Amamojula okukhiqiza atholakala ngenethiwekhi yabasabalalisi: I-T5000 iqala ku-$2.999 ngama-oda angama-1.000 amayunithi noma ngaphezulu futhi i-T4000 isuka ku-$1.999, kuye ngokulungiselelwa.

Kwezezimoto, i-NVIDIA imemezele okuhlukile okuhloselwe ukushayela ukuzimela, nge DRIVE/Thor vula ukubhuka konjiniyela ezohlanganisa amamojula afanayo T5000 kanye T4000. Abakhiqizi bangeza nge amabhodi omama kanye nezinhlelo zokukhiqiza olunikezwa ozakwethu bokuhlanganisa.

Ngokutholakala komhlaba wonke kwehadiwe kanye nesitaki sesoftware esihlanganisiwe, Ukuhlolwa kukhithi kuthuthelwa ngokushesha ekukhiqizeni, ukunciphisa imijikelezo yokuthuthukiswa kanye nokwenza lula ukukala kusuka kuma-prototypes kuya ekusetshenzisweni kwangempela.

Ukuhlanganisa amandla e-AI, ukuxhumana kwezimboni kanye nezindawo ezivuthiwe ze-ecosystem I-Jetson AGX Thor njengenketho eqinile yalabo abadinga amarobhothi akwazi ukubona, ukuqonda, nokwenza ngesikhathi sangempela., kusukela kolayini bemihlangano nasezindaweni zokugcina izimpahla kuya emagumbini okuhlinzela namapulazi.

Indoda ekhubazekile ilawula ingalo yerobhothi-0

I-athikili ehlobene:

Indoda ekhubazekile ilawula ingalo yerobhothi ngengqondo yayo ngenxa ye-interface entsha

U-Alberto Navarro

Ngingumshisekeli wezobuchwepheshe oguqule izintshisekelo zakhe "ze-geek" zaba umsebenzi. Ngichithe iminyaka engaphezu kwengu-10 yempilo yami ngisebenzisa ubuchwepheshe obusezingeni eliphezulu kanye nokukitaza ngazo zonke izinhlobo zezinhlelo ngenxa yelukuluku lokufuna ukwazi. Manje sengiqeqeshelwe ubuchwepheshe be-computer nemidlalo yama-video. Lokhu kungenxa yokuthi sekuphele iminyaka engaphezu kwengu-5 ngisebenza ngokubhalela amawebhusayithi ahlukahlukene ezobuchwepheshe nemidlalo yevidiyo, ngenza izindatshana ezifuna ukukunikeza imininingwane oyidingayo ngolimi oluqondakala yiwo wonke umuntu.

Uma unemibuzo, ulwazi lwami lusukela kuyo yonke into ehlobene nesistimu yokusebenza ye-Windows kanye ne-Android yomakhalekhukhwini. Futhi ukuzibophezela kwami kuwe, ngihlala ngizimisele ukuchitha amaminithi ambalwa futhi ngikusize uxazulule noma yimiphi imibuzo ongase ube nayo kulo mhlaba we-inthanethi.