UNvidia wethula ikhithi lobuhlakani bokufakelwa obusebenzisa Ubuntu

IJetson Nano Developer Kit isebenzisa Ubuntu njengohlelo lokusebenza.

Ikhithi Yonjiniyela yeNVIDIA Jetson Nano kuyikhompyutha eyi-99 dollar eyakhelwe labo abanentshisekelo kumaphrojekthi we-Artificial Intelligence. Yenzelwe ukusetshenziswa ne-Ubuntu 18.04 namathuluzi wemithombo evulekile.

IJetson Nano Developer Kit ihloselwe abakhiqizi bedivayisi abashumekiwe, abacwaningi, nama-hobbyists, okubavumela ukuthi bakhe amakithi abo. Can isetshenziselwe ukusebenzisa izinhlelo zokuhlukaniswa kwesithombe, ukutholwa kwento, ukuhlukaniswa, nokucutshungulwa kwenkulumo.
.
I-NVIDIA Jetson Nano Development Kit iqukethe iJetson chip (engathengwa ngokwehlukana) efakwe kuqala kubhodi yomama. Ikhithi ifaka ukunikezwa kwamandla, izintambo zamandla ezidingekayo, nesoftware efakwe kuqala ekhadini le-MicroSD.

Ikhompyutha ingasetshenziswa ngokubambisana ne-NVIDIA JetPack SDK ekuhlinzeka ngayo imvelo egcwele yedeskithophu ye-Linux ngokusekelwe ku-Ubuntu 18.04 LTS, okubandakanya, ukusheshisa ihluzo, ukuhambisana nethuluzi lamathuluzi le-NVIDIA CUDA kanye nemitapo yolwazi ye-cuDNN 7.3 kanye neTensorRT 5 ″.

Ngokuya nge-NVIDIA, onjiniyela bazokwazi faka kalula izixazululo zomthombo ovulekile ezisebenzayo zokufunda ngomshini (ML), njengeTensorFlow, iCaffe neKeras. Kwenzeka okufanayo naku izinhlaka zokubona ngamakhompyutha nokuthuthukiswa kwamarobhothi njenge-OpenCV ne-ROS.

Izici

I-NVIDIA iqinisekisa ukuthi iJetson Nano ihlinzeka ngama-472 GFLOPS wokusebenza kwekhompyutha kusuka ku-64-bit quad-core ARM CPU yayo kanye ne-128-core NVIDIA GPU

Uhlu oluphelele lwezincazelo zalo ngu:

I-64-bit Quad-core ARM A57 CPU @ 1.43GHz
I-128-core NVIDIA Maxwell GPU @ 921MHz
I-RAM 4GB 64-bit LPDDR4 @ 1600MHz | 25.6 GB / s
I-VIDEO (encoded) 4Kp30 | (4x) 1080p30 | (2x) 1080p60
IVIDIYO (enikwe ikhodi) 4Kp60 | (2x) 4Kp30 | (8x) 1080p30 | (4x) 1080p60
ICAMERA 1x MIPI CSI-2 DPHY imizila
I-RED Gigabit Ethernet, M.2 Ukhiye E
IVIDIYO OUT HDMI 2.0 ne-eDP 1.4

Ukuxhumana kuyimfuneko ebalulekile yalolu hlobo lwedivayisi. Kungakho inezimbobo ezine ze-USB, kanye ne-GPIO, I4C, I2S, SPI ne-UART.

Noma kunjalo, akufanele ukhathazeke ngokusetshenziswa kwamandla.
UJetson Nano kudinga u-5W kuphela wamandla ukusebenzisa (kufika ku-10W kuzinhlelo zokusebenza ezinamandla).

Akuyona ikhompyutha eyodwa yebhodi engasetshenziselwa ukusetshenziswa okuhlukahlukene njenge-Raspberry Pi noma amagajethi wokwakha afana ne-Arduino. Inokusetshenziswa okuqondile, kepha akekho noyedwa owake wathi asikwazi ukuthola izicelo ze-Artificial Intelligence ekhaya lethu.

Angazi ngawe, kepha sengivele ngayiqala incwadi kaSanta Claus.