This is just following Google's Edge TPU, which probably competes with a Raspberry Pi + Movidius stick. The market there is getting interesting.
This is just following Google's Edge TPU, which probably competes with a Raspberry Pi + Movidius stick. The market there is getting interesting.
A lot of third party carrier boards also have a complete sh_tshow of connectors. Auvidea's boards, for example, ship with a Raspberry Pi camera connector, but Raspberry Pi NoIR cameras don't have TX2 drivers, and there are hardly any other cameras that ship with that connector.
Seriously, NVIDIA: Please sell a TX2 devkit that has six non-weird 2-lane CSI connectors and some IMX290 or AR0521 or any other commonly-used robotics sensors for $100 each that plug in and "just work". It would make a lot of people happy to have something to at least start with, and pave the way for third party options to follow the same form factor, connectors, pinouts, board sizes, and so forth.
According to their blog post it actually has driver support for the RPi CM2 8MP (IMX219) and they'll be releasing their own Nvidia-sanctioned cameras available from their partners.
It should hopefully just work. No lowlight options at this time however, which means external CCTV is out of the question :(
I bought a tx2 carrier from connecttech, and half their tx2 boards use a 30 pin connector used by leopard imaging, and the other half use the same ribbon connector that the nvidia devkit uses for its camera. I have $600 worth of camera which doesn't fit the carrier I chose ::face-palm::.
Their xavier carrier, http://connecttech.com/product-category/form-factors/nvidia-..., uses the same connector as the tx2 and xavier dev kits.
Why wouldn't you use the Ethernet port with a traditional GigE camera? I've also done some simple projects using an ordinary USB webcam.
I'm interested in this as a small form factor industrial computer. I've run Raspberry Pis and Intel NUCs in lots of manufacturing equipment where you need something that can run a few lines of Python and sit between the PLC and your device. Given this board's processing power, it might be interesting to plug into a little Dalsa or Basler camera and run a vision algorithm. You can already buy simple "vision sensors" from Keyence, Banner, Sick, etc. that integrate what I understand to be a simple ARM chip with the vision sensor and run basic vision algorithms, but they're often hamstrung by the tooling. The ability to perform and communicate results of arbitrary commands in applications where you don't need lots of processing power would be great.
What camera applications are people builidng that need 1.5 Gbps of camera data? I've built assembly lines that build several parts per second and never even come close to being limited by frame rate or network bandwidth.