Jetson AGX Xavier • Hacker News Board

Re: Jetson AGX Xavier

#21

Post by dchichkov » Sun, May 03, 2020, 8:39 PM UTC

smg wrote: post #13

The cheapest Volta GPUs I have seen so far cost over 2K for 12GB. Can the GPU provided in this kit be used for training?

Yes, if your model is small enough or, if you are fine-tuning small number of layers. TendorFlow 1.15 and 2.0 are available on Xavier. I understand that PyTorch could be built as well.

Nite that the number of CUDA kernels and amount of memory available is smaller, if compared to descrete Volta GPUs.

Re: Jetson AGX Xavier

#22

Post by epmaybe » Sun, May 03, 2020, 8:41 PM UTC

TaylorAlexander wrote: post #11

fvv wrote: post #3

I'm definitely not expert and probably this is a dumb question , but why smart edge things like smart robot and not dumb edge with smart central brain ? Anyway data are useful aggregated central;ly why not incorporate the brain centrally too?

Well, first some clarification - "edge" means "on robot" versus something in the cloud. And the reason you do this is latency and connectivity. I am designing a four wheel drive robot using the NVIDIA AGX Xavier [1] that will follow trails on its own or follow the operator on trails. You don't want your robot to lose cellular coverage and become useless. Even if you had coverage, there would be significant data usage…

This is a bit off topic, but I'm constantly looking at ways to efficiently stream 4K cameras live to local displays as well as remote displays at the highest framerate and resolution possible. How feasible would it be on the xavier to stream 2 4k cameras and display them on at least 2 4k screens? Extra points if you could do that and simultaneously upload to a streaming service, such as twitch.

Re: Jetson AGX Xavier

#23

Post by corysama » Sun, May 03, 2020, 8:44 PM UTC

fvv wrote: post #3

I'm definitely not expert and probably this is a dumb question , but why smart edge things like smart robot and not dumb edge with smart central brain ? Anyway data are useful aggregated central;ly why not incorporate the brain centrally too?

The network is very, very unreliable at the edge. Better to have each piece work independently and store up processed results to transmit eventually, opportunistically. If that processing involves real time video processing there's no way you're going to get that done over a reliably unreliable connection.

Re: Jetson AGX Xavier

#24

Post by fizixer » Sun, May 03, 2020, 8:45 PM UTC

dchichkov wrote: post #21

smg wrote: post #13

The cheapest Volta GPUs I have seen so far cost over 2K for 12GB. Can the GPU provided in this kit be used for training?

Yes, if your model is small enough or, if you are fine-tuning small number of layers. TendorFlow 1.15 and 2.0 are available on Xavier. I understand that PyTorch could be built as well. Nite that the number of CUDA kernels and amount of memory available is smaller, if compared to descrete Volta GPUs.

You say it can do training for small models because of the presence of the small (512-core) GPU? (plus maybe some left-over, control calculations by the CPU)

Re: Jetson AGX Xavier

#25

Post by augustt » Sun, May 03, 2020, 8:47 PM UTC

farseer wrote: post #12

Honest question, can something like this kill the market for embedded DSP processors made by Texas Instruments or Analog Devices?

I mean, this is absurdly more powerful than those dedicated DSPs.

Re: Jetson AGX Xavier

#26

Post by corysama » Sun, May 03, 2020, 8:48 PM UTC

fizixer wrote: post #24

dchichkov wrote: post #21

Earlier quoted context omitted.

Yes, if your model is small enough or, if you are fine-tuning small number of layers. TendorFlow 1.15 and 2.0 are available on Xavier. I understand that PyTorch could be built as well. Nite that the number of CUDA kernels and amount of memory available is smaller, if compared to descrete Volta GPUs.

You say it can do training for small models because of the presence of the small (512-core) GPU? (plus maybe some left-over, control calculations by the CPU)

It's a low-wattage device. It's performance can't hold a candle to a last-gen card that uses 10X the power.

Re: Jetson AGX Xavier

#27

Post by bitwize » Sun, May 03, 2020, 8:51 PM UTC

fvv wrote: post #3

I'm definitely not expert and probably this is a dumb question , but why smart edge things like smart robot and not dumb edge with smart central brain ? Anyway data are useful aggregated central;ly why not incorporate the brain centrally too?

Why do humans carry large, energy-hungry brains around as opposed to being simple tools of the Hivemind like their brethren the insects?

Making the edges smarter allows them to react and adapt on smaller timescales.

Re: Jetson AGX Xavier

#28

Post by my123 » Sun, May 03, 2020, 8:52 PM UTC

weinzierl wrote: post #2

They seem to offer a cheaper 8 GB model too but unfortunately I see no price for it. I'm curious how much it'll be because, as much as I'd like to toy around with this, the $699 is a little to much for just experimentation. EDIT: The 8GB Module seems to be $679 here[1]. This makes the $699 or the 32 GB Developer Kit seem like a steal. Still, too expensive for play, I guess I'll stick with my Jetson Nanos for a while.…

I see that AGX Xavier devkit as much more of an Arm desktop platform, which it is very suitable for too.

Re: Jetson AGX Xavier

#29

Post by my123 » Sun, May 03, 2020, 8:53 PM UTC

fizixer wrote: post #14

Inference only. (So this is competing with Google TPUv1; a few years late and way more expensive, but with more memory)

It can do training with its GPU, not the fastest thing in the world though.

Re: Jetson AGX Xavier

#30

Post by wmf » Sun, May 03, 2020, 8:54 PM UTC

fizixer wrote: post #14

Inference only. (So this is competing with Google TPUv1; a few years late and way more expensive, but with more memory)

You can't put a TPUv1 in a car because Google doesn't sell them.