NVIDIA Announces Jetson AGX Xavier

NVIDIA is pleased to announce Jetson AGX Xavier, the latest addition to the Jetson platform with greater than 10x the energy efficiency and more than 20x the performance of it’s predecessor, Jetson TX2. It’s an AI computer for autonomous machines, delivering the performance of a GPU workstation in an embedded module under 30W. With multiple operating modes at 10W, 15W, and 30W, Jetson Xavier has a peak performance of more than an incredible 30 TOPS (teraops) of mixed-precision FP32/FP16/INT8 performance. Please refer to the FAQ for further information.

Check out our latest blog, and to order see here.

Technical Specs

  • GPU — 512-core Volta GPU with Tensor Cores
  • DL/ML Accelerator — (2x) NVIDIA Deep Learning Accelerator (DLA) Engines (NVDLA.org)
  • CPU — 8-core ARMv8.2 64-bit CPU, 8MB L2 + 4MB L3
  • Memory — 16GB 256-bit LPDDR4x | 137 GB/s
  • Storage — 32GB eMMC 5.1
  • Vision Accelerator — 7-way VLIW processor
  • Video Encode — (2x) 4Kp60 | HEVC
  • Video Decode — (2x) 4Kp60 | 12-bit support
  • Mechanical — 100mm x 87mm with 16mm Z-height                            (699-pin board-to-board connector)

I/O

  • Display — 3x eDP/DP/HDMI at 4Kp60 | HDMI 2.0, DP HBR3
  • Camera — 16x CSI-2 Lanes (40 Gbps in D-PHY V1.2 or 109 GBps in CPHY v1.1)                     8x SLVS-EC lanes (up to 18.4 Gbps)                     Up to 36 virtual channels
  • PCIe — 5x 16GT/s gen4 controllers | 1x8, 1x4, 1x2, 2x1                  (3x) Root Port + Endpoint                  (2x) Root Port
  • USB — (3x) USB 3.1 (10GT/s)               (4x) USB 2.0 Ports
  • Ethernet — Gigabit Ethernet-AVB over RGMII
  • Other I/Os — UFS, I2S, I2C, SPI, CAN, GPIO, UART, SD

Jetson Xavier Developer Kit

The Jetson Xavier Developer Kit is currently able to be ordered from the NVIDIA webstore in the following countries:

  • US, Canada, UK, France, Germany, Italy, Poland, Spain, and China

Additionally, orders can be placed through distributors for:

  • Japan, Europe, Israel, and Turkey (other regions to follow soon)

Initial orders will be fulfilled on a first-come first-served basis within an estimated six-week time period beginning early September 2018.
The price of the devkit for Registered Developers is $1,299 (limit 1 unit per Registered Developer, prices vary by region). Anyone can qualify for the special price by registering an account at developer.nvidia.com. Additional units are available in quantities 1-10 at $2,499 and quantities >10 are $1,599.

Documentation & Resources

FAQ — https://developer.nvidia.com/embedded/jetson-xavier-faq
Blog — https://news.developer.nvidia.com/nvidia-jetson-agx_xavier-developer-kit-now-available/
Order — https://developer.nvidia.com/embedded/buy/jetson-xavier-devkit?ncid=so-int-jnxrpr-57255

First post! :D I’m looking forward to this machine. I feel like I have to ask some questions to post, so I came up with some.

Will the Isaac SDK be released simultaneously with Xavier or on a seperate schedule? Also, I wonder what robots are available for simulation. I see these so I suspect they are included in Isaac Sim,

Why will it be available to order?
Thanks!

Hi Ramsey, welcome to the forums, the Isaac availability will follow the Jetson Xavier availability by a few weeks next month.

Hi guo.tang, it should be quite soon! Make sure you sign up for Notify Me on this page:
https://developer.nvidia.com/jetson-xavier-devkit

Hi!, i got an email from nvidia for an 24-hour early access… But the problem is i live in japan and the only option i got when pre-ordering is US, CANADA only…

Hi soliva, Japan pre-orders should be enabled soon, next week timeframe if you check back.

Hello.
I want to pre-ordering.
But why is $2499? Not should $1299?
I want to buy 6 or more about it.
Please help me thank you very much!

Hi horatiolui, Registered Developers can order a unit at $1299 (anyone can register at developer.nvidia.com).
For quantities 2-10, the price is $2499, and for >10 units the price is $1599.

Keep refreshing the page in hope of a UK order option! On a separate note, any idea when the wiki will be up again?

@morten.hanasand

Page now up -

Hi Morten, I will be creating and populating the new eLinux wiki for Jetson Xavier shortly before shipments begin.

I think Morton was referring to the error message on the wiki -

"MediaWiki 1.31 internal error

MediaWiki 1.31 requires at least PHP version 7.0.0 or HHVM version 3.18.5, you are using PHP 5.6.37.
Supported PHP versions

Please consider upgrading your copy of PHP. PHP versions less than 5.6.0 are no longer supported by the PHP Group and will not receive security or bugfix updates.

If for some reason you are unable to upgrade your PHP version, you will need to download an older version of MediaWiki from our website. See our compatibility page for details of which versions are compatible with prior versions of PHP."

Awesome, cheers!

@dusty_nv

Yes, issue with the php server, I was looking for some SPI info for the tx2 carrier, but just found what I needed in one of the pdfs, so no worries.

Thanks for letting me know, we don’t host the eLinux wiki server ourselves so not sure what happened - I will reach out to it’s admin right now. Cheers.

From checking with the admin, it sounds like they have moved servers, and that the wiki should be back up later on today.

OK, the eLinux server migration has completed and they are back up. Thanks Morten and jjhadsworth!

NVIDIA is now accepting pre-orders for the Jetson Xavier Developer Kit! For the latest info, see our recent blog.

Pre-orders are currently able to be placed in North America, Europe, Japan, and China. Other regions will follow soon.

Initial pre-orders will be fulfilled on a first-come first-served basis within an estimated six-week time period beginning early September 2018.

To pre-order the devkit, see here — https://developer.nvidia.com/embedded/buy/jetson-xavier-devkit?ncid=so-int-jnxrpr-57255

Great, thanks for looking into it Dusty¡

Hi :)
whats the microarchitecture of the ARM cores? Cortex-A55, Cortex-A75, Cortex-A77 or an internal implementation?
Is it a big.LITTLE architecture or just big?

Edit:
Does it support ARM CoreSight?