Nvidia Jetson(s) Explained - in under 400 seconds!
HTML-код
- Опубликовано: 27 сен 2024
- Confused which #nvidia Jetson board to choose? Watch this video for a quick #overview of all the Jetson series, starting from #jetson TK1 to the newest (just introduced on #gtc 2024!) Jetson Thor - plus my own personal comment on the best board choice.
Links to development kits (NO affiliate links!):
Rugged Industrial Orin Nano 4 Gb - www.seeedstudi...
Cheapest Orin Nano 4 Gb DevKit - www.waveshare....
Official Orin Nano 8 Gb DevKit - www.seeedstudi...
Currently Cheapest Jetson Nano 4 Gb DevKit - www.waveshare....
Resources:
www.nvidia.com...
developer.nvid...
developer.nvid...
blogs.nvidia.c...
developer.nvid...
developer.nvid...
Support my work on making tutorials and guides on Patreon!
www.patreon.com/hardware_ai
I can't say it 100% relieved my confusion, but it did provide a lot of food for thought! ty
Glad it helped!
Good info. Thanks.
Glad it was helpful!
I still don't understand that SOM part? Can I start with it like it is cheaper so can I save my money by starting to build the robot with it? Or should I buy the developer kit first? I have no experiences with jetsons though...
You can't use SoM directly - unless you have a suitable carrier board for it :) dev kit is SoM + carrier board.
For hobby stuff, you just get a dev kit (official or third party one, third party ones might be cheaper or have extra capabilities).
For production, you first develop software / test things with a dev kit, while creating a custom carrier board for the SoM that will be in your product.
so this can't be used for ML , DL or Stable difusion , can it?
Well, there is more than one board in the video. Which one are you asking about?
@@Hardwareaigosh just answer his question!! Show all kit used those
@@bosscs that is a truckload of attitude for a person watching a free RUclips video, that took me 10+ hours to make....
Regardless, for the question itself, it's just not very properly formulated, so I wanted to get more input.
ML, DL - all of the boards can be used for running some sort of ML inference.
Stable Diffusion - there are different models out there, some are smaller, some are bigger. There are some tutorials on this topic specifically in Jetson AI lab
www.jetson-ai-lab.com/tutorial_stable-diffusion.html
White David Lee Frank Lee Susan
Bot, beep-beep.
Jetson Nano 0.476TOPS 😑
Jetson Xavier 25-40TOPs (Interersting)
Jetson Orin 200TOPs (Nice but expensive)
Jetson Thor 0.8PETAFLOPS (A little monster, start controling humanoids 2024)
Next Gen over Thor 4PETAFLOPS (Robots take over the Industry 2026)
Human Brain 12PETAFLOPS
Next Gen Over Next Gen 20PFLOPS (Robots take over the World 2028)
Make Earth Great Again 2028 :) or MEGA
In what application do you lose by not having NVLDA in the Orin Nano development kit like the Orin Nano NX has?
Mostly when you need to run another model or another instance of the same model simultaneously. forums.developer.nvidia.com/t/i-dont-know-the-dla/81869/2