I’ve ordered two Zed 2i cameras for my thesis. I plan to calibrate the cameras and merge the point clouds to get a large scale map.
Yet, I’ve not found a paper about hardware performance.
Which hardware would be the most interesting between:
- Jetson XT1
- Jetson Nano
- Jetson Xavier
- Intel NUC I7 10th gen with 64GB RAM
Thanks for your help.
The ZED SDK requires CUDA to run, so you need a device with an NVIDIA GPU.
For embedded applications the NVIDIA Jetson devices are the right choice, I suggest taking in consideration at least a Jetson Xavier NX with a fast SSD memory to save the big amount of generated data in real-time.
For more computational power, NVIDIA AGX Orin is the best available solution.
You can find ZED Box with Xavier NX and NVIDIA AGX Orin in our online store:
Thank you very much for your quick answer.
I’ll go with the Xavier as I can easily pick one from my university.
If it’s not enough powerful for my project, I’ll try to buy a AGX Orin.