Ibhodi yophuhliso ArduinoKwiinguqu zayo ezahlukeneyo, lukhetho olukhulu kubenzi kunye nabaphuhlisi abafuna ukwenza ezabo iiprojekthi ze-DIY. Kwelinye icala nawe unayo Raspberry Pi, I-SBC encinci nexabiso eliphantsi yokwenza uninzi lweeprojekthi. Kuko konke oku kongezwe inani elikhulu kakhulu lezinye iindlela kwintengiso. Kodwa mhlawumbi ujonge into ethe ngqo, into oza kusebenza nayo uthungelwano lwe-neural kunye ne-AI. Ke ibhodi yeNVIDIA Jetson Nano yile nto uyifunayo.
Ukuba ucwangcisa ukuphuhlisa okanye ukufunda ngamanethiwekhi e-neural, ungathembela kule bhodi I-NVIDIA Jetson Nano ngayo. Konke ngaphezulu nje kwe- € 100, hayi phezulu kakhulu xa kuthathelwa ingqalelo amaxabiso ezinye iinkqubo zobuhlakani ...
Yintoni uJetson?
I-NVIDIA Jetson Nano yiprojekthi yenkampani eyaziwayo yemizobo ye-chip eyenza ukuba kuphuculwe iinkqubo ezininzi ezincinci ze-AI. Ukongeza, ukujonga amaxabiso ezinye iiprojekthi zothungelwano lwe-neural, ikwenza oko ngexabiso kunye nokusetyenziswa kwamandla okuphantsi.
Ngale bhodi yophuhliso unokwenza ubuninzi beeprojekthi, ezinje ngezicelo ze-IoT, ezivela kwiirobhothi ezincinci zasekhaya, ukuya kwezinye iinkqubo ezikrelekrele onokucinga ngazo, ngokusebenzisa iirekhodi zevidiyo zenethiwekhi (i-NVRs), amasango akrelekrele, njl.
Konke kwimodyuli encinci yePCB emalunga ne-70x45mm ngobukhulu kuguqulelo lwayo lweNano, eyona incinci. Ngapha koko, yi Uhlobo lwe-SOM yemodyuli (Inkqubo yeModyuli) ikulungele ukusebenza nayo, kunye nekhithi yophuhliso.
Iimpawu zobugcisa
Ewe iimpawu zobugcisa Ukusuka kwi-NVIDIA Jetson Nano, uza kuzifumanela ibhodi enamandla ekwaziyo ukuphuhlisa ii-GFLOPs ezingama-472 zomsebenzi wokuphumeza ii-algorithms ze-AI zale mihla ngokukhawuleza. Inokuqhuba nothungelwano oluninzi lwe-neural ngokudibeneyo kunye nenkqubo ye-sensors eziphakamileyo ezininzi ngaxeshanye.
Kwaye konke oko kusetyenziswa ngokufanelekileyo amandla. Ngaphandle kwamandla ayo, kuphela Idla phakathi kwe-5 kunye ne-10w. Amandla asezantsi ngokwentelekiso ejonga izibonelelo ezibonelelwayo.
Ngolwazi oluthe kratya, ndikushiyela itafile epheleleyo ye iinkcukacha:
| GPU | Uyilo loyilo lweNVIDIA Maxwell™ nge-128 NVIDIA CUDA cores® | |
| ICPU | Inkqubo ye-ARM® Cortex®-A57 MPCore I-Quad Core | |
| Memoria | 4GB i-4-bit LPDDR64 | |
| Ukugcina | 16GB eMMC 5.1 Ukugcinwa kweflash | |
| Ukufakwa kwikhowudi kwevidiyo | Iifayile ezingama-4K ezingama-30 (H.264 / H.265) | |
| Ukucaciswa kwevidiyo | Iifayile ezingama-4K ezingama-60 (H.264 / H.265) | |
| Ikhamera | Indlela eyi-12 (3 x 4 okanye 4 x 2) MIPI CSI-2 DPHY 1.1 (18 Gbps) | |
| Conectividad | Gigabit Ethernet | |
| Isikrini | HDMI 2.0 okanye DP 1.2 | eDP 1.4 | I-DSI (1 x 2) 2 ngaxeshanye | |
| KUKHULU | 1 1/2/4 PCIE, 1 USB 3.0, 3 i-USB 2.0 | |
| NGU | 1 I-SDIO / 2 SPI / 4 I2C / 2 I2S / GPIO | |
| Tamaño | 69,6 mm x 45 mm | |
| Oomatshini | Ikhonkco engama-260 | |
Ezinye iimveliso ezifanayo
I-NVIDIA ikwanikezela ezinye iimveliso ezifanayo kuJetson Nano wophuhliso lwe-AI kunye neenethiwekhi ezingezizo ezenziweyo. Eminye imizekelo yile:
- UJetson Xavier NXImodyuli ye-SOM enika onke amandla ekhompyuter kunye nemilinganiselo emincinci kakhulu. Unokunyuka ukuya kwii-TOP ezingama-21, oko kukuthi, ukusebenza kweTera okungama-21 ngemizuzwana. Ngaphezulu kwamandla awaneleyo okuqhuba iinethiwekhi zangoku ze-neural ngokudibeneyo kunye nenkqubo yedatha evela kwiimvakalelo ezininzi zokulungisa ngexesha elinye.
- UJetson AGX XavierImodyuli entsha ephawula inqanaba ngokubaluleka koxinano nokusebenza ngokufanelekileyo. Kwi-AI, evumela ukwenziwa kwezizukulwana ezitsha zoomatshini abakrelekrele.
- IJetson TX2-Enye isantya esiphezulu, ibhodi yophuhliso olunamandla okusebenza kwekhompyuter ye-AI. Ikhompyuter ephezulu kwimodyuli esekwe kwi-NVIDIA Pascal GPU. Ukuya kuthi ga kwi-8GB ye-RAM kunye ne-bandwidth ye-59,7GB / s.
Ewe, njengoko ubona, abantakwabo abadala xabiso iphezulu kakhulu ...
Fumana i-NVIDIA Jetson Nano
Ukuba uthatha isigqibo soku thenga iNVIDIA Jetson Nano, unamathuba amaninzi kuyo. Enye yeemveliso ezinikezelwa ngeqonga leAmarzon. Uyakufumana zombini ibhodi yophuhliso ngaphandle kokungaphezulu, okanye iikiti eziphuculweyo ezipheleleyo ezibandakanya iadaptha yamandla, njl. Umzekelo:
- I-NVIDIA Jetson Nano esisiseko
- I-NVIDIA Jetson Nano Kit ene-adaptha yamandla, i-64GB microSD, i-USB
- Thenga kuphela imodyuli ye-SOM
Khumbula ukuba ubuchwepheshe be yokufunda umatshini, ubukrelekrele bokuzenzela, ukufunda okunzulu, njl. Ke ngoko, ukufunda ngazo kunokuba nomdla ekwenzeni iiprojekthi ezintsha zexesha elizayo okanye ufumane imisebenzi enomdla kwiinkampani ezifuna olu hlobo lwezakhono ...