Connect Tech Carrier Boards and Jetson™ Modules
EMPOWERING

Edge AI Hardware and Software Solutions
Industry-leading end-to-end Edge AI solutions from a single trusted partner and a Global Leader for NVIDIA® Jetson Platforms
COTS - Prototyping - SWaP
THE JUNCTION is a distributor of professional data connectivity and IT infrastructure products and components. We provide reliable, high quality computer accessories that enable your devices. Our suppliers rank amongst the world’s technology leaders in each of their fields.
We support for both legacy and standard PCs, PC/104 derivatives, Small Form Factor technologies, CompactPCI, OpenVPX, and others. The areas of deployment cover the entire gamut of commercial and industrial applications. These range from relatively straightforward serial I/O wired connections to complex industrial and maritime applications. These include PLC automation, alarm systems, medical journaling, industrial flow control, rail and road signalling, voice recording, PBX, sewerage, oil & gas, UAVs, AUVs, ROVs, ATC, Space Research, aeronautics, and more.
Our portfolio includes both standard commercial and industrial products designed for use in harsh conditions that demand a wide temperature range and, in certain cases, consideration for vibration.
Several of our engineering partners offer custom design facilities. Among such clients to whom we have delivered equipment in recent years are SAAB (devices for monitoring UAVs and other aircraft as part of the EADS MIDCAS project), Kongsberg Maritime and Kapsch TrafficCom.
Connect Tech is a hardware design and manufacturing company that specializes in rugged, small form factor solutions. They have a nimble engineering team ready to engage in Custom Design when “off-the-shelf” is not an option. Innovative design is not all they do – the company is driven by the belief that custom satisfaction is as important as a quality product. It is their goal to continue this tradition in the future development of innovative designs and to exceed the expectations of all customers with whom they come into contact.
As an industry-leading provider of NVIDIA© Jetson™ hardware, Connect Tech is constantly striving to develop adjacent technologies to meet the growing needs of the Edge Computing community. Their Carrier Boards and Embedded devices support the full line of Jetson modules, including the NVIDIA© Jetson™ Nano, TX2, TX2i, Xavier NX, AGX Xavier, AGX Orin. Orin NX, Orin Nano and AGX Thor.
AUTHORIZED DISTRIBUTOR / RESELLER
Center4
+47 40 40 34 02
Info@Junction.eu
www.TheJunction.eu
PUBLISHER
The Junction AS
CONCEPT AND DESIGN
Digital Office s.r.l. www.design-office.ro
ARTWORK
www.shutterstock.com
www.connecttech.com www.nivida.com
Connect Tech is happy to support the new members of the Jetson family, NVIDIA Jetson Thor™ , the AGX Orin™, Orin™ Nano and the Orin™ NX, with considerable expertise supporting the NVIDIA Jetson AGX Xavier™, Jetson™ TX2/TX2i/TX1, and Jetson Nano™ platforms.
To complement the Jetson Thor™ and Orin™, Connect Tech has released a series of Carrier boards, as well as thermal cooling solutions and adapters.
Connect Tech also provides custom design services for Thor™ and Orin™ users who require a unique feature set as well as a quick time to market strategy.
If you're looking for a Jetson Carrier board designer with extensive experience, then Connect Tech is the perfect partner for your unique design needs!
Among the Jetson modules supported are the NVIDIA Jetson Thor™ , NVIDIA AGX Orin™, Orin™ Nano, Orin™ NX Jetson Nano™ , Jetson™ TX2/TX2i/TX1, Xavier NX, and Xavier™ .
Connect Tech's team of Jetson specialists will work with your team to produce a custom solution that perfectly matches your criteria while assuring product performance and reliability, from minor adjustments to full custom design.
Their engineering design services team excels at producing unique solutions that meet both time-to-market and cost objectives.
AGX201 Forge & AGX202 Rogue
Use the most advanced artificial intelligence computers available to create autonomous machines that are both low in energy consumption and environmentally friendly. High-speed interface support for multiple sensors, along with up to 275 Tflops of processing power and 8 times the capability of the previous generation for multiple concurrent AI inference pipelines, make NVIDIA® Jetson Orin™ modules the right choice for the next generation of robots.
Check out our most recent NVIDIA AGX Orin and Orin NX offerings – see the Table of Contents or refer to the part number in the Part Number Overview:
Want to learn more?
Click here
Astro - Jetson TX2i/TX2/TX2-4GB
Boson for FRAMOS Carrier
Cogswell - Jetson TX2/TX2i
Elroy - Jetson TX2i/TX2
Forge - Jetson AGX Orin™
Gauntlet Carrier for Jetson Thor™
Hadron Carrier for Jetson Orin™ NX
Hadron-DM Carrier for NVIDIA® Jetson Orin™ NX
Super Hadron-DM Carrier for Jetson Orin™ NX and Nano
Orbitty - Jetson TX2i/TX2 Carrier
• OrbittyBox
Photon - Jetson AI Camera Platform
Quark - Jetson Xavier™ NX & Nano
Quasar - Jetson TX2/TX2i
Rogue Carrier - Jetson AGX Xavier™
Rogue Carrier for Jetson AGX Orin™
Rogue-RX - Jetson AGX Orin™
Spacely - Jetson TX2i/TX2
Sprocket Carrier for Jetson TX2/TX2i
GMSL2 Camera Board
• e-con Systems | Intel® RealSense™ | Leopard Imaging
Allied Vision MIPI Camera Board
FPD-Link III Camera Board
NVIDIA® Jetson™ SDI Vision Platform
AGX111
AGX201
AGX202
AGX203
AGX301
ASG001
ASG002
ASG003
ASG006
ASG007
ASG008
ASG009
ASG012
ASG016
ASG021
ASG026
ASG029
CBG341
D7G001
JCB002
JCB003-01
JCB005
JCB006
JCB010 NGX002
https://connecttech.com/resource-center/l4t-board-support-packages/
• PoE PD (NGX002) capable, power via separate input or over Ethernet (IEEE 802.3af-2003 and IEEE 802.3at2009 compatible)
• 1x GbE
• 1x NVMe (M.2 M-Key)
• 1x microSD
• 4x GPIO, I2C
• SB Console/UART
• USB OTG for programming
• DC barrel power input also available
• Dimensions: 145 x 64,5 mm
Connect Tech’s Photon is a Jetson AI Camera Platform specifically designed to support smart camera applications. This Jetson platform provides PoE PD Gigabit Ethernet uplink for data and power and has various I/O connectors positioned for easy packaging into off the shelf camera housings. For applications that don’t require PoE PD, ask for CTI’s cost effective non-PoE PD options. Let Connect Tech take your Jetson Nano™ or Jetson Xavier™ NX solution to a commercially deployable platform today.
While Photon is compatible with NVIDIA Jetson Nano, TX2 NX, and Xavier NX SOMs, some I/O availability will change depending on the module. For a complete overview, see the Photon Compatibility Overview.
Integrations with external storage and WiFi/Bluetooth via Intel are also available Contact us for further information
Note: Do not use the 19V power supply from the Jetson Xavier™ NX Dev Kit with the Photon.
NGX002 - Photon - Jetson AI Camera Platform multi camera input, GbE, PoE uplink
NGX002-81
NGX003 - Photon - Jetson AI Camera Platform multi camera input, GbE, no PoE uplink
Not recommended for new designs. Please see Hadron-DM Carrier Board.
The Orin™ NX Hadron Carrier Board is an ultra-compact, rugged, and feature-rich carrier for AI Computing at the Edge.
Hadron, which is only slightly larger than the Jetson™ SODIMM module, is ideal for vision applications, inference, and unmanned payloads. Users can switch between modules based on their AI processing needs because the Hadron carrier board is compatible with the Jetson Orin™ NX, Orin™ Nano, Xavier™ NX, and TX2™ NX modules.
Hadron's design features a wide input voltage range, a robust I/O set, and locking IO connectors optimized for harsh environments.
• Compatible with Jetson Orin™ NX, Orin™ Nano
• Ideal for Vision Applications, Inference and unmanned payloads
• Rugged Latching connectors
• 1x 1000BASE-T Port
• 2x USB3.1
• 1x 4-Lane MIPI CSI2, 22-pin FPC Connector
• 1x 2242/2230 NVMe M.2
• Wide Power Input: +9V to +60V DC
• Display: None - Headless Operation only
• Operating Temp.: -25 to +85 C
• Weight: 49g
• Dimensions: 82,6 x 58,8 mm
NGX012 - Hadron Carrier for NVIDIA® Jetson Orin™ NX Part No. Jetson
Cooling
NGX012-07 Orin NX 8GB X 1TB
NGX012-09 Orin NX 8GB X 1TB
NGX012-10 Orin NX 8GB X WiFi5 1TB
NGX012-25 Orin NX 16GB X 1TB
NGX012-26 Orin NX 16GB X 1TB
NGX012-27 Orin NX 16GB X 1TB
NGX012-28 Orin NX 16GB X WiFi5 1TB
NGX012-29 Orin NX 16GB X WiFi5 1TB
NGX012-30 Orin NX 16GB X WiFi5 1TB
NGX012 - Recommended optional accessories
Active Thermal Passive Heatsink for the NVIDIA® Jetson Xavier™ NX Module XHG313 Thermal plate for NVIDIA® Jetson Xavier™ NX Module Cabling
CBG112 Power Cable - 6-pin MiniTek w/ Latch, unterminated
CBG117 Ethernet Cable – 10-pin MiniTek w/ Latch, RJ-45 Panel Mount
CBG136 CR2032 RTC Battery w/ 3-pin Connector Cable Assembly 70 mm - Molex P/N 6032101013
CBG615 USB2.0 Cable - 20-pin Tiger-Eye to 2x USB 2.0 Type-A Female
CBG629 I/O cable - 40-pin Tiger-Eye to Unterminated Flying Leads - SFSD-20-28C-G-12.00-SR
CBG686
Power
MSG070 Astro/Photon PSU - AC/DC Wall Mount (Europe). 90-264VAC Input, 12VDC Output, 2A Max Output, 24W NVIDIA® Jetson Orin™ NX Modules
900-13767-0010-000 Jetson Orin NX 8GB
900-13767-0000-001 Jetson Orin NX 16GB
• 1x USB 3.1, 2x GbE,
• 2x 2-lane MIPI CSI-2, 1x USB OTG
• 1x SD card slot, 3x 3.3V UART, 2 x I2C, 1x CAN 2.0b, and 1x SPI
• +5V DC Input
• Smaller than the Dev Kit: 82,6 x 58,8 mm
NGX004 -Quark Carrier for NVIDIA® Jetson Nano™
Cooling options
NGX014 - Recommended optional accessories Part No.
XHG308 NVIDIA® Jetson Nano™ Passive Heat Sink
XHG309 NVIDIA® Jetson Nano™ Active Heat Sink
XHG310 NVIDIA® Jetson Nano™ Thermal Transfer Plate
CBG112 Power Cable - 6-pin MiniTek w/ Latch, unterminated
CBG117 Ethernet Cable – 10-pin MiniTek w/ Latch, RJ-45 Panel Mount
CBG125 I/O Cable – DF20A 40 Pin, unterminated
CBG136 CR2032 RTC Battery w/ 3-pin Connector Cable Assembly 70 mm - Molex P/N 6032101013
MSG070 Astro/Photon PSU - AC/DC Wall Mount (Europe). 90-264VAC Input, 12VDC Output, 2A Max Output, 24W NVIDIA® Jetson Orin™ Nano Modules
900-13767-0040-000 Jetson Orin Nano 4GB 900-13767-0030-000 Jetson Orin Nano 8GB
** For additional information about NGX014, please refer to PCN #: CTIU-00045.
Connect Tech's Quark Carrier for Edge AI computing is feature-rich, compact, and affordable. It is somewhat larger than the Jetson SODIMM module and ideal for unmanned payloads, inference, and vision. They are interoperable with Jetson Nano, TX2 NX, and Xavier NX SoMs, so users can switch modules if their processing needs vary.
Positive-locking Quark offers 1x USB 3.1, 2x GbE, 2x 2-lane MIPI CSI-2, 1x USB OTG, 1x SD card slot, 3x 3.3V UART, 2x I2C, and 1x SPI, as well as toughenvironment IO connectors.
Quark supports NVIDIA Jetson Nano, TX2 NX, and Xavier NX SoMs, but module I/O availability may vary.
For a detailed explanation, see the Quark Compatibility Specifications.
Note: Do not use the 19V power supply from the Jetson Xavier™ NX Dev Kit with the Quark Carrier.
• Integrate up to 4x 2-lane or 3x 4-lane MIPI FRAMOS Sensor Modules
• Dual Gigabit Ethernet and Dual USB
• NVMe and microSD card for additional storage
• WiFi and Bluetooth expansion options
• +9V to +36V wide input voltage range
AI vision powerhouse Boson for FRAMOS integrates up to four MIPI cameras in a tiny size. Boson optimizes sensor inputs and storage for high-end vision applications in the FRAMOS Sensor Ecosystem. This collaborative product offers a plug-and-play vision solution that speeds up application development. The Jetson Nano, TX2 NX, and Xavier NX SoMs allow customers to switch modules as their processing needs vary.
NGX017 - Boson for FRAMOS Carrier Board for NVIDIA® Jetson Xavier™ NX Part No. Jetson module
Thermal Cooling options
WiFi5/BT NVMe Thermal Plate Passive Active
NGX017
NGX017-11 Xavier NX 8GB X X
NGX017-36 Xavier NX 16GB X
Boson for FRAMOS's board support package includes software hooks that instantly connect various FRAMOS Sensor Module Ecosystem cameras to NVIDIA's JetPack SDK, eliminating your software team's development needs.
Boson contains four FRAMOS PixelMate connectors on the carrier board. These connectors let users easily connect 4x 2-lane or 3x 4-lane MIPI CSI-2 cameras from the FRAMOS Sensor Module Ecosystem.
Boson for FRAMOS supports NVIDIA Jetson Nano, TX2 NX, and Xavier NX SoMs, but I/O availability varies. See the Boson for FRAMOS Compatibility Specification Table for details.
For differencies between the NGX007 and the NGX017, please refer to PCN (Product Change Notice) at this URL: https://connecttech.com/ftp/pdf/CTIU-00046.pdf
Assembly Instructions fro CBG136: https://connecttech.com/ftp/pdf/cables/CBG136_CTIC-00477.pdf
NGX017 - Recommended optional accessories
Part No.
XHG311 Passive Finned Passive Heatsink for the NVIDIA® Jetson Xavier™ NX Module
XHG312 Active Thermal Passive Heatsink for the NVIDIA® Jetson Xavier™ NX Module
XHG313 Thermal plate for NVIDIA® Jetson Xavier™ NX Module Cabling
CBG125 I/O Cable – DF20A 40 Pin, unterminated
CBG136 CR2032 RTC Battery w/ 3-pin Connector Cable Assembly 70 mm - Molex P/N 6032101013
CBG408 Power Input Cable
CBG509 Camera I/O Cable Power Adapter
MSG085 External Power Supply (Class 1) for Rogue/Rudi-AGX/Rudi-NX - 19V 120W AC/DC. Includes CBG315 Input Power Cable and CBG411 4-pin DIN to 4-Pin Mini-Fit Jr. NVIDIA® Jetson Xavier™ NX Modules
900-83668-0000-000 Jetson Xavier NX 8GB
900-83668-0030-000 Jetson Xavier NX 16GB
• Compatible with Jetson Orin™ NX and Orin™ Nano
• Rugged Latching Connectors
• Tiny footprint: 82,7mm x 68,8mm
• 2x USB3.1, 1x GbE, 2x GMSL Coax Ports
• 1x NVMe M-Key, 1x M.2 E-Key (WiFi/BT)
• Wide Power Input: +9V to +60V DC
With two GMSL2 camera inputs, the Hadron GMSL Carrier Board for Orin™ NX and Orin™ Nano is a very sturdy, feature-rich, and ultra-compact solution. It guarantees dependable, fast data transmission, which is essential for independent applications. Hadron is a great option for AI computing at the Edge because it is just marginally bigger than the Jetson™ SODIMM module and is excellent for GMSL vision applications, inference processing, and unmanned payloads. Hadron's architecture has locking I/O connectors that are ideal for harsh settings, a complete I/O set, and the ability to take a wide input voltage range.
Hadron Carrier Comparison Chart
NGX018 - Recommended optional accessories
Part No.
XHG323 Orin NX / Nano - Thermal Transfer Plate
XHG324 Orin NX / Nano - Passive Heatsink
XHG325 Orin NX / Nano - Active Heatsink Cabling
CBG117 Ethernet Cable – 10-pin MiniTek w/ Latch, RJ-45 Panel Mount
CBG136 CR2032 RTC Battery w/ 3-pin Connector Cable Assembly 70 mm - Molex P/N 6032101013
CBG615 USB2.0 Cable - 20-pin Tiger-Eye to 2x USB 2.0 Type-A Female
CBG619 GMSL Coax Cable - Samtec MMCX Plug to FAKRA Plug Z-Type IP67
CBG629 I/O cable - 40-pin Tiger-Eye to Unterminated Flying Leads - SFSD-20-28C-G-12.00-SR
CBG686 USB3.0 Cable - 20-pin Tiger-Eye to 2x USB 3.0 Type-A Female (200 mm)
CBG705 Force Recovery and Reset - Molex 2 Position PicoBlade Connector to Panel Mount Pushbutton
CBG706 Power Cable - Molex Nano-Fit 2x2 Off-the-Shelf Cable Assembly Molex Part Number: 451300410 Breakout Board
XBG023 USB 3.0 Breakout Board - 20-pin Tiger-Eye to 2x USB 3.0 Type-A Female Power Adapter
MSG113 Power Supply GMSL & Dual MIPI: NGX018, NGX024, NGX025 NVIDIA® Jetson Orin™ NX Modules
900-13767-0010-000 Jetson Orin NX 8GB
900-13767-0000-001 Jetson Orin NX 16GB
• Compatible with NVIDIA® Jetson Orin™ Nano, and Orin™ NX.
• Integrate up to 3x 2-lane or 2x 4-lane MIPI FRAMOS Sensor Modules
• Dual Gigabit Ethernet and Dual USB
• NVMe for additional storage
• WiFi and Bluetooth expansion options
• +9V to +36V wide input voltage range
NGX020 - Boson (FRAMOS) Carrier Board for NVIDIA® Jetson Orin™ NX & Orin™ Nano
The AI vision powerhouse Boson can fit up to four MIPI cameras into its remarkably compact design. With Boson, high-end vision applications can make the most of their sensor inputs and storage solutions. This carrier board shortens the time it takes to develop applications by giving consumers a vision solution that is plug-and-play. If your processing demands change, you may easily switch between modules on your NVIDIA® Jetson OrinTM Nano and OrinTM NX.
The carrier board of the Boson contains four FRAMOS PixelMate connectors. Users are able to directly connect a maximum of two 4-lane or four 2-lane MIPI CSI-2 cameras with these connectors.
Alternatives for Boson Integration
A lower total hardware stack height is achieved by directly integrating the camera to the carrier board, rather than using an adapter board. If a customer wants a complete stack solution, they can link Boson to their preferred NVIDIA module, any accessible heat sinks, and more.
NGX020 - Recommended optional accessories
XHG323 Orin NX / Nano - Thermal Transfer Plate
XHG324 Orin NX / Nano - Passive Heatsink
XHG325 Orin NX / Nano - Active Heatsink
CBG125 I/O Cable – DF20A 40 Pin, unterminated
CBG408 Power Input Cable
CBG509 Camera I/O Cable
900-13767-0010-000 Jetson Orin NX 8GB
900-13767-0000-001 Jetson Orin NX 16GB
900-13767-0040-000 Jetson Orin Nano 4GB
900-13767-0030-000 Jetson Orin Nano 8GB
• MiPi Camera Inputs:
o 2x 4-lane MiPi Modules or o 4x 2-lane MiPi Modules or
o 2 x 4-lane and 2x 2-lane MiPi Modules
Dual Gigabit Ethernet, USB-C, USB 3.0 NVMe for additional storage; WiFi and Bluetooth expansion options
• +9V to +36V wide input voltage range
• 22pin MiPi CS12 connectors for additional camera compatibility
NGX021 - Boson-22 Carrier Board for NVIDIA® Jetson Orin™ NX & Orin™ Nano
options
An AI vision powerhouse, Boson-22 for Jetson OrinTM NX integrates up to four MIPI cameras into an incredibly compact design.
Optimizing sensor inputs and storage solutions for high-end vision applications is the goal of Boson-22 for Jetson OrinTM NX. Development of apps desiring to employ several cameras on the Jetson platform will be accelerated by Boson-22, which is compatible with numerous sensor modules and equipped with PC style interfaces.
A lower total hardware stack height is achieved by directly integrating the camera to the carrier board, rather than using an adapter board. If a customer wants a complete stack solution, they can link Boson to their preferred NVIDIA module, any accessible heat sinks, and more
NGX021 - Recommended optional accessories
Part No. Description Thermal Solutions
XHG323 Orin NX / Nano - Thermal Transfer Plate
XHG324 Orin NX / Nano - Passive Heatsink
XHG325 Orin NX / Nano - Active Heatsink Cabling
CBG125 I/O Cable – DF20A 40 Pin, unterminated
CBG408 Power Input Cable
CBG509 Camera I/O Cable
Jetson Orin
900-13767-0010-000 Jetson Orin NX 8GB
900-13767-0000-001 Jetson Orin NX 16GB
900-13767-0040-000 Jetson Orin Nano 4GB
900-13767-0030-000 Jetson Orin Nano 8GB
• 4x FPD Link III, ISB 3.0, USB 2.0, CAN 2.0b, USB OTG, RS-485, I2C, GPIO, SPI, PWM
• 1x NVMe (PCIe x4, 2280)
• Operating Temp.: -20 to +80C
• WiFi/BT expansion options
• Passive, Active and Thermal Transfer Plate cooling options
Specifically designed for the NVIDIA® Jetson OrinTM NX and Jetson OrinTM Nano modules, this Lepton FPD-Link III Carrier Board was created with the explicit intention of ensuring smooth integration and providing additional capabilities. Lepton is outfitted with a total of four USB 3.0 connectors and two GbE connections, each of which offers a substantial amount of USB connectivity. As a result of its four FPD-Link III camera inputs, NVMe storage expansion connectivity, and flexible M.2 slots architecture, Lepton offers versatility that may be utilized for a wide variety of applications.
Part No.
NGX022 - Recommended optional accessories
Description
Thermal Solutions
XHG323 Orin NX / Nano - Thermal Transfer Plate
XHG324 Orin NX / Nano - Passive Heatsink
XHG325 Orin NX / Nano - Active Heatsink Cabling
CBG408 Power Input Cable
CBG341 Quad FAKRA GMSL Cable length 1m
CBG629 I/O cable - 40-pin Tiger-Eye to Unterminated Flying Leads - SFSD-20-28C-G-12.00-SR
CBG814 CAN Cable - SFSD-03-28C-G-12.00-SR
Power Adapter
MSG085 External Power Supply (Class 1) for Rogue/Rudi-AGX/Rudi-NX - 19V 120W AC/DC.
Includes CBG315 Input Power Cable and CBG411 4-pin DIN to 4-Pin Mini-Fit Jr.
NVIDIA® Jetson Orin™ NX & Orin™ Nano Modules
900-13767-0010-000 Jetson Orin NX 8GB
900-13767-0000-001 Jetson Orin NX 16GB
900-13767-0040-000 Jetson Orin Nano 4GB
900-13767-0030-000 Jetson Orin Nano 8GB
• Compatible with NVIDIA® Jetson Orin™ Nano, and Orin™ NX.
• Ideal for Vision Applications, Inference and unmanned payloads
• Rugged Latching connectors
• 2x 4-Lane MIPI CSI2, 22-pin FPC Connector
• 2 x USB 3.1. 1x GbE, 1 x 2242/2230 NVMe, 1x M.2 E.Key (WiFi/BT)
• GPIO, PwM, i2C, RS-232, 3,3V UART, SPI
• Tiny footprint: 82,7mm x 68,8mm
• +9V to +36V wide input voltage range
The Hadron-DM Carrier Board for Jetson Orin™ NX is a compact, durable, and feature-rich carrier for AI computing at the edge. Hadron, slightly larger than the Jetson™ SODIMM module, is suitable for vision applications, inference, and unmanned payloads. Hadron-DM has dual MIPI CSI-2 connectors for increased adaptability. The Hadron-DM carrier board is compatible with the Jetson Orin™ NX and Jetson Orin™ Nano modules, allowing users to switch between modules based on their AI processing needs.
NGX024 - Hadron-DM Carrier for NVIDIA® Jetson Orin™ NX & Orin™ Nano Part No. Jetson module Thermal Cooling options WiFi5/BT NVMe Thermal Plate Passive Active
NGX024
NGX024-07 Orin NX 8GB X 1TB
NGX024-09 Orin NX 8GB X 1TB
NGX024-19 Orin NX 16GB X 0,5TB
NGX024-25 Orin NX 16GB X 1TB
NGX024-26 Orin NX 16GB X 1TB
NGX024-27 Orin NX 16GB X 1TB
NGX024-30 Orin NX 16GB X WiFi5 1TB
NGX024 - Recommended optional accessories
Part No. Description Thermal Solutions
XHG323 Orin NX / Nano - Thermal Transfer Plate
XHG324 Orin NX / Nano - Passive Heatsink
XHG325 Orin NX / Nano - Active Heatsink Cabling
CBG112 Power Cable - 6-pin MiniTek w/ Latch, unterminated
CBG117 Ethernet Cable – 10-pin MiniTek w/ Latch, RJ-45 Panel Mount
CBG136 CR2032 RTC Battery w/ 3-pin Connector Cable Assembly 70 mm - Molex P/N 6032101013
CBG615 USB2.0 Cable - 20-pin Tiger-Eye to 2x USB 2.0 Type-A Female
CBG629 I/O cable - 40-pin Tiger-Eye to Unterminated Flying Leads - SFSD-20-28C-G-12.00-SR
CBG686 USB3.0 Cable - 20-pin Tiger-Eye to 2x USB 3.0 Type-A Female (200 mm)
CBG706 Power Cable - Molex Nano-Fit 2x2 Off-the-Shelf Cable Assembly Molex Part Number: 451300410
CBG732 Power Cable - Nano-Fit 2x2 to DC Jack Breakout Board
XBG023 USB 3.0 Breakout Board - 20-pin Tiger-Eye to 2x USB 3.0 Type-A Female Power Adapter
MSG113 Power Supply GMSL & Dual MIPI: NGX018, NGX024, NGX025 NVIDIA® Jetson Orin™ NX Modules
900-13767-0010-000 Jetson Orin NX 8GB
900-13767-0000-001 Jetson Orin NX 16GB
900-13767-0040-000 Jetson Orin Nano 4GB
900-13767-0030-000 Jetson Orin Nano 8GB
The Super Hadron-DM Carrier Board for Jetson Orin™ Nano is a compact, durable, and feature-rich carrier for AI computing at the edge. It is slightly larger than the Jetson™ SODIMM module, and is suitable for vision applications, inference, and unmanned payloads. The Super Hadron-DM has Dual MIPI CSI-2 connectors for increased adaptability.
The Jetson Orin™ NX and Jetson Orin™ Nano modules provide Super Mode, allowing users to switch between modules based on AI processing needs.
Hadron Carrier Comparison Chart
NGX027 - Super Hadron-DM Carrier for NVIDIA® Jetson Orin™ NX & Orin™ Nano
NGX027 - Recommended optional accessories
Part No. Description Thermal Solutions
XHG323 Orin NX / Nano - Thermal Transfer Plate
XHG324 Orin NX / Nano - Passive Heatsink
XHG325 Orin NX / Nano - Active Heatsink Cabling
CBG112 Power Cable - 6-pin MiniTek w/ Latch, unterminated
CBG117 Ethernet Cable – 10-pin MiniTek w/ Latch, RJ-45 Panel Mount
CBG136 CR2032 RTC Battery w/ 3-pin Connector Cable Assembly 70 mm - Molex P/N 6032101013
CBG615 USB2.0 Cable - 20-pin Tiger-Eye to 2x USB 2.0 Type-A Female
CBG629 I/O cable - 40-pin Tiger-Eye to Unterminated Flying Leads - SFSD-20-28C-G-12.00-SR
CBG686 USB3.0 Cable - 20-pin Tiger-Eye to 2x USB 3.0 Type-A Female (200 mm)
CBG706 Power Cable - Molex Nano-Fit 2x2 Off-the-Shelf Cable Assembly Molex Part Number: 451300410
CBG732 Power Cable - Nano-Fit 2x2 to DC Jack Breakout Board
XBG023 USB 3.0 Breakout Board - 20-pin Tiger-Eye to 2x USB 3.0 Type-A Female Power Adapter
MSG113 Power Supply GMSL & Dual MIPI: NGX018, NGX024, NGX025 NVIDIA® Jetson Orin™ NX Modules
900-13767-0010-000 Jetson Orin NX 8GB
900-13767-0000-001 Jetson Orin NX 16GB
900-13767-0040-000 Jetson Orin Nano 4GB
900-13767-0030-000 Jetson Orin Nano 8GB
• 6x 2-lane or 4x 4-lane MIPI CSI Camera Inputs
• 2x NVMe M.2 Key M Slots, 3x USB 3.1, 2x GbE
• Wide input power range 9-19V DC
• Commercially deployable NVIDIA® Jetson AGX Xavier™ platform
• Dimensions: 92 x 105 mm
Connect Tech’s Rogue is specifically designed for commercially deployable platforms. It is a full featured Carrier Board for the NVIDIA® Jetson AGX Xavier™ module with an extremely small footprint of 92 x 105 mm.
The Rogue adds 3x USB 3.1, 2x GbE, 2x HDMI, and a locking Mini-Fit Jr. power input connector to an amazing range of latest generation interfaces on the Jetson AGX Xavier.
A selection of Rugged camera add-on expansion boards is available for use with the Rogue to link directly with the Jetson AGX Xavier high density MIPI CSI ports. Integrations with external storage and WiFi/Bluetooth via Intel 8265 are also available, contact sales@connecttech.com for more details.
To understand the differences between the AGX101 and the AGX111, please consult the PCN to be found at this address: https://connecttech.com/resource-center/pcn-ctiu-00040-nvidia-jetson-agx-xavier-products/
AGX111
AGX111-01 AGX Xavier 32GB X
AGX111-07 AGX Xavier 32GB X
AGX111-08 AGX Xavier 32GB X 1TB
AGX111-12 AGX Xavier 32GB X WiFi5 2TB
AGX111-13 AGX Xavier 32GB X
AGX111-14 AGX Xavier 32GB X
AGX111-17 AGX Xavier 32GB X
AGX111-25 AGX Xavier 32GB X
AGX111-31 AGX Xavier 32GB X
AGX111-37 AGX Xavier 32GB X
AGX111-61 AGX Xavier 32GB X JCB002
AGX111-62 AGX Xavier 32GB X 1TB JCB002
AGX111-69 AGX Xavier 32GB X 2TB JCB002
AGX111-77 AGX Xavier 32GB X WiFi5
AGX111-83 AGX Xavier 32GB
AGX111-M5 AGX Xavier 64GB X
AGX111-N1 AGX Xavier 64GB X
AGX111-N2 AGX Xavier 64GB X 1TB
Recommended Optional Components
JCB002
JCB005
• Commercially deployable NVIDIA® Jetson AGX Orin™ platform
• 1x 16-lane MIPI Expansion Connector
• 2x NVMe M.2 Key M Slots, 2x USB 3.2, 2x GbE
• Wide input power range 10-36V DC
• 1792-core NVIDIA Ampere architecture GPU with 56 Tensor Cores
• 8-core Arm® Cortex®-A78AE v8.2 64-bit CPU 2MB L2 + 4MB L3
• 32GB 256-bit LPDDRS 204,8 GB/s
• 64GB eMMC 5.1
• Dimensions: 155 x 125 mm
Connect Forge is a full-featured Carrier Board for the NVIDIA® Jetson AGX Orin™ module, specifically designed for commercially deployable platforms. On the Jetson AGX Orin, Forge provides access to an impressive list of latest generation interfaces while also adding 3x USB 3.2, 2x GbE, Display Port, and a locking 6-pin Mini-Fit Jr. power input connector. It also supports rugged camera add-on expansion boards that interface directly with high density MIPI CSI interfaces. External storage and WiFi/Bluetooth via Intel 8265 integrations are also available; please contact us for more information.
AGX201 - Forge Carrier Board for NVIDIA® Jetson AGX Orin™
AGX201-001 AGX Orin 32GB X AGX201-004 AGX Orin 32GB X
AGX201-09D AGX Orin 32GB X
AGX201-0CD AGX Orin 64GB X
AGX Orin 64GB X
AGX Orin 64GB X
AGX Orin 64GB
AGX Orin 64GB
AGX Orin 64GB
AGX Orin 64GB X
AGX201-0MT AGX
64GB
AGX201-133 AGX Orin 64GB X JCB002 AGX201-1CD AGX Orin 64GB X 1TB JCB002
AGX201-1QT AGX Orin 32GB X 2TB JCB005
AGX201-232 AGX Orin 64GB X JCB005
AGX201-2TS AGX Orin 64GB X JCB003
AGX201-367 AGX Orin 32GB
AGX201-397 AGX Orin 64GB X JCB003
AGX201-398 AGX Orin 64GB X
JCB003
AGX201-3FD AGX Orin 64GB X JCB003
AGX201-3JG AGX Orin 64GB X JCB003
Video Encode
2x8K @30 | 6x4K @60 | 12x4K @30 | 26x1080p @60 | 52x1080p @30 (HEVC) | 30x1080p @30 (H.264) 1x 4K60 | 3x 4K30 | 6x 1080p60 | 12x 1080p30 (H.265) H.264, AV1
Video Decode (2x) 8Kp30 / (6x) 4Kp60
18x 1080p30 (H.265) H.264, VP9, AV1
|
|
1080p30 (H.265) H.264, AV1
|
8K30 | 3x 4K60 | 7x 4K30 | 11x 1080p60| 22x 1080p30 (H.265) H.264, VP9, AV1
+ 1x4 + 1x2 + 2x1
Rogue Carrier for Jetson AGX Orin™ is a full-featured carrier board designed specifically for commercially deployable platforms.
Rogue for Jetson AGX Orin™ gives you access to the Jetson AGX Orin's impressive list of latest generation interfaces while also adding 2x 10GbE, 3x USB 3.1, HDMI, and a locking Mini-Fit Jr. power input connector.
Integrations with external storage and WiFi/Bluetooth via Intel 8265 are also available; contact Sales@Center4.eu for further information.
Features include:
• Compatible with NVIDIA Jetson AGX Orin
• Dual 10GBASE-T (10G Ethernet)
• Camera inputs: All connected via high-density camera connector breakout. Same Pinout as development kit. Application specific camera add-on boards can be used in conjunction with this interface (Up to 6x MIPI I-PEX, SerDes Inputs: GMSL or FPD-Link III, HDMI Inputs)
• Operating Temp.: -40 to +85 C
• Dimensions: 92 x 107 mm
- Rogue Carrier Board for NVIDIA® Jetson AGX Orin™
Note: * For the Passive Heatsink external airflow required Recommended Optional Components
AGX202-001 AGX Orin 32GB X
AGX Orin 64GB X
AGX Orin 64GB X
AGX Orin 32GB
AGX202-1CJ AGX Orin 64GB
AGX202-1TU AGX Orin 64GB X
AGX202-232 AGX Orin 64GB X JCB005
AGX202-23B AGX Orin 64GB X WiFi5 2
AGX202-2TT AGX Orin 64GB
AGX Orin 64GB
AGX202-3MQ AGX Orin Ind. 64GB X WiFi5 AGX202-3TS AGX Orin Ind. 64GB X 1TB
AGX202-3WV AGX Orin Ind. 64GB X 1TB
AGX202-3ZX AGX Orin Ind. 64GB X JCB002
AGX202-4MK AGX Orin Ind. 64GB X 1TB JCB003
• Commercially deployable NVIDIA® Jetson AGX Orin™ and AGX Orin™ Industrial platforms
• 2x 10GBASE-T w/Rugged Positive Locking Connectors
• 2x NVMe PCIe Gen4 M.2 Key-M 2280 Slots
• Dimensions: 92mm x 108.2mm
The fully functional Rogue-RX Carrier for Jetson AGX Orin™ was designed especially for hardy and unforgiving conditions. The Rogue-RX Carrier for Jetson AGX Orin™ offers a distinct wire/pin header with positive locking for UART, I2C, SPI, CAN, and GPIO in addition to robust positive locking high-speed connections for 10GBASE-T, USB 3.2 Gen2, and HDMI.
Please email sales@connecttech.com for more information about the WiFi/Bluetooth and external storage integrations that are also available using the Intel 8265.
AGX203 - Rogue-RX Carrier Board for NVIDIA® Jetson AGX Orin™
Note: * For the Passive Heatsink external airflow required
Recommended Optional Components
AGX203-0CE AGX Orin 64GB X
AGX203-0JR AGX Orin 64GB X
AGX203-0MN AGX Orin 64GB X
AGX203-3MK AGX Orin 64GB Ind. X
AGX203-3WU AGX Orin 64GB Ind. X
AGX203-3WV AGX Orin 64GB Ind. X 1TB
AGX Orin 64GB Ind. X
AGX203-498 AGX Orin 64GB Ind. X
JCB002 AGX203-4JG AGX Orin 64GB Ind. X
Specifications
Compatibility NVIDIA® Jetson Thor™
Networking
2x 10GBASE-T (10G Ethernet)
2x 1000BASE-T (1G Ethernet)
Display 1x DisplayPort
Camera Inputs
USB
Storage
I2C/SPI
UART
1x 16-Lane MIPI Expansion Connector
Interface directly to a range of GMSL3, GMSL2, FPD-Link III, SDI, and MIPI CSI-2 image sensors
1x USB 3.2 Ports (Type-C – OTG Capable), 1x USB 3.2 Ports (Type-C)
2x M.2 Key M 2280 NVMe (1x 4 Lane PCIe Gen3, 1x 2 Lane PCIe Gen 3)
2x I2C Channel @ 3.3V IO
2x SPI Channel @ 3.3V IO
2x RS232/422/485
2x @3.3V levels UART
CAN 4x CAN 2.0b Non-Isolated Ports
4x 3.3V GPI Inputs
GPIO
User Expansion
Input Power
6x 3.3V GPO Outputs (2x PWM Capable)
1x 3.3V Power Pin at 1A
1x M.2 Key E 2230 WiFi/BT (PCIe/USB2)
1x M.2 Key B 3042/3052 (LTE/5G – USB3) with Nano SIM
+30V to +60V DC Wide Input Power (6-Pin Mini-Fit Jr. Connector)
Dimensions 155mm x 126mm
Operating Temp. -40C to +85C
Connect Tech's Gauntlet Carrier Board, designed for the NVIDIA® Jetson Thor™ Series Modules, offers industrial-grade performance, ultra-fast networking, and advanced sensor integration. It accelerates the next generation of physical AI, robotics, autonomy, and edge AI.
Gauntlet maximises the performance of NVIDIA's Blackwell GPU by providing a broad set of interfaces and expansion options.
Gauntlet is fully interoperable with Connect Tech's vision-ready platforms, allowing developers to rapidly create full-stack AI systems with a variety of imaging needs.
Compact. Rugged. Deployable. Gauntlet reflects the future of robots.
The NVIDIA® Jetson Thor™ Series Modules are the ultimate platform for physical AI and robotics, delivering up to 2070 FP4/1035 FP8 TFLOPs with Blackwell GPU and 128GB RAM. With a 14-core Neoverse ARM CPU and 4X25 GbE networking for real-time multi-sensor processing, it elevates your projects and integrates generative AI. Combine with Connect Tech solutions to fully utilise the Jetson Thor™ SoM's processing power and sensor interface capabilities for edge AI applications.
Connect Tech’s GMSL camera platform is an expansion board that allows up to 8 cameras to be connected to the Jetson AGX Xavier module. Since many different types of GMSL cameras are available, the JCB002 has user selectable options to interface with either GMSL1 or GMSL2 protocols at different operating frequencies. Power to the cameras is provided by PoC (Power over Coax) so all the data, control signals, and power are sent through a single 50 Ohm Coaxial cable. This allows for flexibility in cable routing and ease of installation in automotive applications.
The main power for the board comes from the Camera Expansion Header. The 12V power for the cameras is also available from the same header on the Rogue carrier. NVIDIA’s development kit does not provide voltage here, so a hot pluggable external 12V connector is available on the JCB002. The circuit will always take the external voltage if both are connected.
• Compact size: 75 x 57 mm
• GMSL1 or GMSL2 protocols
• Internal or External Camera power
• Power over Coax
• 4x mini coax connector
• 4%input voltage protection
• Software power cycle
GMSL camera platform is an expansion board that allows up to 8 cameras to be connected to the Jetson AGX Xavier module. Many different types of GMSL cameras are available; the JCB002 has user selectable options to interface with either GMSL1 or GMSL2 protocols at different operating frequencies.
NVIDIA Xavier Connection (Uplink)
(GMSL) Camera Inputs
De-serializer
MIPI Output
Camera Input Connectors
Board Dimension:
1x High Density Connector
Camera Board will mate to the NVIDIA Jetson “Camera Expansion Header”
8x Total (GMSL2/GMSL1)
Maxim MAX9296A
A single 4-lane MIPI CSI-2 v1.3 output from each De-serializer (16-lanes total)
2x MATE-AX Quad Coax Connectors
Breakout cables to FAKRA available
75 x 57 mm
PoC (Power-Over-COAX) All 8 cameras will be sourced Power-Over-COAX from JCB002
Power JCB002 can be directly powered from Camera Expansion Header or External +12V Input
Operating Temperature -40 C to +85 C
Weight
Warranty and Support
50g
1 Year Warranty and Free Support
Currently supported cameras
As Edge AI Vision's capabilities expand, discovering ways to readily combine sensors with embedded hardware enables rapid product development and field deployment. As NVIDIA's largest Jetson hardware partner, we've collaborated with industry-leading vision camera and sensor manufacturers to create preintegrated and validated solutions. Customers can reduce total application development time by using reliable hardware that includes the appropriate camera and sensor software.
JCB003-01
Please keep in mind that the Jetson SDI Vision Platform is only supported on Connect Tech's Rogue carrier boards for NVIDIA® Jetson AGX Xavier™, as well as the Rogue and Forge Carrier Boards for NVIDIA® Jetson AGX Orin™ .
• 2x 3G-SDI/HD-SDI/SDI (via 2x HD-BNC Connectors) SDI Video Inputs
• 1x 3G-SDI/HD-SDI/SDI (via 1x HD-BNC Connector) SDI Video Outputs
• PCBA Dimensions: 67mm x 75mm PCBA, Height with Integrated passive heatsink: 17mm
• 53 grams
• SDI Video Inputs 2x 3G-SDI/HD-SDI/SDI (via 2x HD-BNC Connectors)
• Video Input Path to Jetson SDI-to-MIPI (via Samtec QTH Connector)
• SDI Video Output 1x 3G-SDI/HD-SDI/SDI (via 1x HD-BNC Connector)
• Video Output Path from Jetson HDMI-to-SDI (via HDMI Type-A Connector)
• SDI Supported Formats SMPTE 259M: SD-SDI
• SMPTE 292M: HD-SDI
• SMPTE 424M: 3G-SDI
• MIPI CSI-2 Details Each SDI Input channel will be converted to a 4-lane MIPI CSI-2 v1.3 output per SDI Input (8-lanes total)
• Power +12V Provided via Camera Expansion Connector
• Mechanical PCBA Dimensions: 67mm x 75mm PCBA Height with Integrated passive heatsink: 17mm
• Thermal Design Power 6W
• Operating Temperature -40°C to +85°C
• Weight 53g
• Warranty and Support 1 Year Warranty and Free Support
The NVIDIA® JETSON™ SDI VISION PLATFORM allows for ultra-low hardware latency SDI-to-MIPI CSI-2 conversion, providing a direct ISP ingest path to GPU compute on the NVIDIA® Jetson™ platform without the overhead from other methods such as frame grabbers, USB, or Ethernet packetizing. Ideal for Edge AI applications in robotics, industrial and public sector, supporting multi-sensor, multi-spectral, electro-optic and infrared (EO/IR) surveillance and targeting systems for ISR and Inference applications across the air, land and maritime domains.
• Simple integration of Allied Vision MIPI CSI-2 sensors to the Jetson AGX Xavier platform
• Supports Allied Vision Alvium 1500 C and 1800 C sensors
• Connect up to 6x 2-lane or 4x 4-lane MIPI Cameras
• Seamless integration to the Rogue and Rogue-X Carrier Boards
• Dimensions: 75 x 40,2 mm
Please note Allied Vision MIPI Camera Board is only supported on our Rogue carrier boards for NIVIDIA® AGX Xavier™ and Forge carrier boards for NVIDIA® AGX Orin™
Developers of Jetson AGX Xavier projects can finally rest easy knowing that the Allied Vision MIPI Camera Board provides the vision solution they've been hoping for. This Camera Board enables the easy attachment of up to six MIPI sensors with no additional hardware, allowing for the speedy incorporation of your preferred cameras.
This Camera Board is preconfigured with the device-tree software needed to incorporate up to six Allied Vision Alvium 1500 C and 1800 C sensors into your application, sparing you the hassle of writing custom code. By integrating with the Connect Tech Rogue and Rogue-X Carrier Boards, users can go straight from prototyping to mass production with no need for new software. Allied Vision's Alvium 1500 C and 1800 C camera series both offer MIPI CSI-2 options, with users able to choose between integrating 6x 2-lane or 4x 4-lane cameras.
Not for use with NVIDIA Xavier Dev Kit.
Browse Allied Vision Alvium 1500 C Camera Series – Click Here
Browse Allied Vision Alvium 1800 C Camera Series – Click Here
Additional Board Features
• The extended temperature range of -40 to +85 degrees Celsius makes it ideal for use in industrial settings.
• Capable of drawing +12V power from the Rogue and Rogue-X Carrier's Camera Expansion header
• 8x FPD-Link III camera inputs, 2x per deserializer
• Internal or External Camera power
• Power over Coax
• Input voltage protection
• Dimensions: 75 x 57 mm
Please note FPD-Link III Camera Board is only supported on our Rogue carrier boards for NIVIDIA® AGX Xavier™ and Forge carrier boards for NVIDIA® AGX Orin™
The JCB006 camera expansion board enables the connection of FPD-Link III Deserializers to Rogue or Rogue-X carriers.
PoC (Power over Coax) provides power to the cameras, so all data, control signals, and power are transmitted over a single 50 Ohm Coaxial cable. This allows for greater cable routing flexibility and ease of installation in automotive applications. The Camera Expansion Header provides the board with its primary power.
• Compatible with Forge, Rogue
• Support for up to 4 Ultra HDMI inputs converting each HDMI input to two lanes MiPi CS12
• HDMI to CSI2 conversion for vision processing
• Compatible and tested with the Sony FCB line of block cameras
• Inputs: 2x 30pin coax (Sony FCB compliant), 2x Micro HDMI
• 12V via High Density Connector or external power connector
• Dimensions: 75 mm 52 mm
• Temp. Range: -30 C to +70 C
An HDMI signal can be entered directly into the Connect Tech HDMI Camera platform, which then converts it to MIPI CSI2 so it can be used with the NVIDIA Jetson. With its two KEL USL0030L connectors and two micro HDMI ports, this HDMI receiver is compatible with a broad range of HDMI cameras and other devices.
Astro is specifically designed to work with the NVIDIA® Jetson TX2/TX2i supercomputer-on-module.
The Astro Carrier provides access to features found on the Jetson TX2/TX2i module in an extremely small footprint (87 x 57 mm). It makes use of a high density, board-to-board connector that allows for use with Connect Tech’s off the shelf or custom designed break out boards.
• 2xGbE (10/100/1000) Ports
• 2xUSB2, 1xUSB3 and 1xHDMI Ports
• 1xI2C, 1xAux, 4xGPIO, 1xmini PCIe, 1xmSATA
• Use with COTS or custom break out boards
• Extended Temperature Range -40 C to +85 C (Astro Carrier)
• Video Inputs: MIPI, GMSL
• Dimensions: 87 x 57 mm
Astro Carrier for NVIDIA® Jetson TX2i/TX2/TX2 4-GB
In addition to the common features for the Astro this board offers:
• 3 x CSI
Astro Breakout Board for NVIDIA® Jetson TX2
In addition to the common features for the Astro this board offers:
• 2 x CSI
Astro Breakout Board for NVIDIA® Jetson TX2
ASG012
Astro Breakout Board for NVIDIA® Jetson TX2i/TX2-4GB
ASG012 Astro Carrier Board Integrated Solution powered by NVIDIA® Jetson™ TX2i
ASG012-01 The package consists of Astro Carrier Board (ASG012) integrated with NVIDIA® Jetson™ TX2i (900-83489-0000-000) with CTI-L4T BSP loaded (CMG508) + GMSL populated + Mimic (ASG001) + Breakout Board (XBG206) + Active Thermal (XHG302)
• Head-to-Head Dual Mini-PCIe
• Dual half size or one single full length Mini-PCIe/mSATA expansion
• Dual x2 MIPI CSI-2 Video Input Connectors
• HDMI Video
• USB 3.0 and 2.0
• 1 x GbE
• 2 x RS-232/485
• 1 x microSD, SPI, I2C, GPIO
• Extended Temperature Range -40 C to +85 C (Elroy Carrier)
• Dimensions 87 x 50 mm
Designed to match the NVIDIA® Jetson TX2 and TX2i module form factor, Connect Tech’s Elroy Carrier for NVIDIA® Jetson TX2/TX2i brings a low cost deployable Jetson TX2/TX2i Solution to the market.
Specialized for use in a small form factor rugged environment, with locking pin-header connectors, solder in standoffs, and industrial temperature range components, the Elroy is really going places.
Connect Tech’s Orbitty Carrier for NVIDIA® Jetson TX2, TX2i 4GB, TX2i is designed to match the NVIDIA© Jetson TX2/TX2i module form factor. The
Orbitty is fully equipped with 1x USB 3.0, 1x USB 2.0 OTG, 1x HDMI, 1x GbE, 1x microSD, 2x 3.3V UART, I2C, and 4x GPIO. Ideal for robotics and unmanned systems, as well as any small form factor rugged environment.
Orbitty completes Connect Tech's Jetson TX2 and TX2i product lines by providing our most affordable carrier yet. Get your Jetson™ TX2 module and Orbitty up and running quickly.
• 1x GbE, USB 3.0, USB 2.0, 1x HDMI, 1x MicroSD, 2x 3.3V UART, I2C, 4x GPIO
• +9V to +14V DC Nominal (+19V Peak)
• Extended Temperature Range: -40 C to +85 C (Orbitty Carrier)
• Dimensions: 87 x 50 mm
ASG003
ASG003 Orbitty Carrier Board Integrated Solution
Orbitty Carrier with NVIDIA® Jetson™ TX2 Flashed and Integrated (CMG503-21)
• Specifically designed for use with Jetson TX2 and Orbitty Carrier
• 2-Piece Metal Enclosure
• Weight: Enclosure + Fasteners – 113g
• (Orbitty, Module, Heatsink – 204g)
• Optional 2x SMA Antenna Connectors
• Dimensions WxHxD: 95,2 x 63 x 58.2 mm
Connect Tech’s OrbittyBox easily turns the Orbitty carrier into a complete packaged NVIDIA® Jetson TX2 system.
This two-piece metal enclosure is designed to house an Orbitty Carrier, an NVIDIA® Jetson TX2 module, and Connect Tech's active Heatsink. The enclosure is sold as an accessory, allowing it to be integrated with existing customer products. Its two-piece design allows for simple installation and assembly.
Connect Tech’s Spacely Carrier for NVIDIA® Jetson TX2/TX2i is an ideal product for unmanned vehicle applications, or any application where situational awareness is critical. Spacely enables users to simultaneously connect up to 6 MIPI CSI-2 cameras, as well as offering built-in expansion for a GPS/GNSS module. It includes a multi-I/O port specifically designed to allow easy connection to OEM Autopilots
Common features for all Spacely Carriers:
• Connect up to 6 MIPI CSI-2 Camera Inputs
• Tailored IO for easy connection to Pixhawk Autopilot
• 2x GbE, 1x uHDMI, 2x USB 3.0, 2x USB 2.0, 1x USB OTG
• 1x miniPCIe Slot, mSATA Slot, CAN 2.0b port
• 2x UART, I2C, SPI, 16 GPIO all at +3.3V IO
• Extended Temperature Range -40 C to +85 C (Spacely Carrier)
ASG006 / ASG026
Spacely Carrier for TX2 up to 6 MIPI CSI-2 optional GPS/GNSS module.
ASG009 / ASG029
Spacely Carrier for TX2 up to 6 MIPI CSI with NEO-M8L GPS/GNSS module w/3D sensors
https://connecttech.com/ftp/pdf/CTIU-00067.pdf
• 5 x Gigabit Ethernet Channels – 4 x PoE, 2 x PoE+ PSE Gigabit Ports
• Only a single +12V input required; No need for external 48V PoE power
• 1 x USB 3.0, 1x USB 2.0, 1x USB OTG, 2x RS-232, 1x miniPCIe, 1x mSATA
• Extended Temperature Range: -20 C to +80 C
• Mounting option (Surface Edge Mount)
Specifically designed for use with GigE Vision Cameras. Allowing up to 5 Gigabit Ethernet cameras to be connected, 4 of which can be powered by on board Power over Ethernet. This system is ideal for Machine Vision applications
Accessories
Connect Tech’s Cogswell Carrier for NVIDIA® Jetson TX2 and TX2i is ideal for use in Gigabit Ethernet Vision applications. This product provides Gigabit Ethernet channels with built-in Power over Ethernet (PoE) sourcing capabilities, ideal for use with GigE Vision cameras.
• Slimmest design possible in Z-axis – All components fit “under” TX2/TX2i module
• Extremely Small Size: 87 x 50 mm
• 1x USB OTG, 1 x4 lane MIPI CSI-2, 2x 3.3V UART, I2C, 4x GPIO
• +12V to +16V DC Input Range
• Extended Temperature Range -40 C to +85 C (Sprocket Carrier)
Connect Tech’s Sprocket Carrier for NVIDIA® Jetson TX2/TX2i is designed to match the NVIDIA Jetson module form factor. This design provides our smallest and lowest profile solution yet. The Sprocket and Jetson TX2/TX2i stack can be mounted flat in the smallest of spaces, perfect for space constrained payloads such as that of a drone. Jump right into your next project with your Jetson TX2/TX2i module and Sprocket.
• 1x CAN, 1x 4-lane MIPI (I-PEX), 1x SATA
• 1x GbE, 2x USB 3.0, USB 2.0 OTG, 1x HDMI,
• 1x MicroSD, 2x 3.3V UART, I2C, 4x GPIO
• +9V – +14V DC Nominal (+19V Peak)
• Extended Temperature Range -40 C to +85 C (Quasar Carrier)
• Dimensions: 87 x 50 mm
Connect Tech’s Quasar Carrier for NVIDIA® Jetson TX2/TX2 4GB/TX2i is an update to the Orbitty Carrier, designed to match the NVIDIA Jetson TX2/TX2 4GB/TX2i module form factor. The Quasar design includes all the same connectors as the Orbitty (1x GbE, 1x microSD, 2x 3.3V UART, I2C, and 4x GPIO,1x USB 2.0 OTG), with the addition of 2x USB 3.0 ports, 1x 4lane MIPI (I-PEX), 1x CAN, an updated 1x HDMI 2.0 connector as well as a SATA 7-pin connector for storage while maintaining the extremely small 87 x 50 mm footprint.
Like the Orbitty, the Quasar is ideal for robotics, unmanned applications, or any small form factor rugged environment that require the addition of MIPI, CAN, SATA or HDMI 2.0 port. Quasar is the latest addition to the Connect Tech Jetson TX2, TX2 4GB and TX2i line of products offering a wide variety of connectors in a very small and affordable package.
Get up and running quickly with your Jetson module and Quasar.
Connect Tech offers a variety of Heatsinks and thermal plates for each of its carrier boards. A carrier board requires no more than one. They are screwmounted between the carrier board and the NVIDIA Jetson module. It is important to choose the right Heatsink or thermal plate for the application.
A customer would order a thermal plate when designing their own thermal solution, e.g. if they were planning to mount the module on the wall of an enclosure, which would allow them to use a much larger surface area for heat dissipation. Heat plates cannot be used with an active or passive Heatsink.
XHG301 NVIDIA® Jetson™ TX2 4GB/TX2i Passive Heatsink
XHG302 NVIDIA® Jetson™ TX2 4GB/TX2i Active Heatsink
XHG303 NVIDIA® Jetson™ TX2 4GB/TX2i Passive Heatsink
XHG304 NVIDIA® Jetson™ TX2 4GB/TX2i Active Heatsink
XHG305 NVIDIA® Jetson AGX Xavier™ Passive Heatsink
XHG306 NVIDIA® Jetson AGX Xavier™ Active Heatsink
XHG307 Liquid Cooling Block for NVIDIA Jetson AGX Xavier
XHG307-01 Liquid Cooling Development Kit
XHG308 NVIDIA® Jetson Nano™ Passive Heatsink
XHG309 NVIDIA® Jetson Nano™ Active Heatsink
XHG310 NVIDIA® Jetson Nano™ Thermal Transfer Plate
XHG311 NVIDIA® Jetson Xavier™ NX Passive Heatsink
XHG312 NVIDIA® Jetson Xavier™ NX Active Heatsink
XHG313 NVIDIA® Jetson Xavier™ NX Thermal Transfer Plate
XHG314 NVIDIA® Jetson Xavier™ NX Active Heatsink
XHG316 NVIDIA® Jetson™ TX2 NX Thermal Transfer Plate
XHG317 NVIDIA® Jetson™ TX2 NX Passive Heatsink
XHG318 NVIDIA® Jetson™ TX2 NX ™ Active Heatsink
XHG319 NVIDIA® Jetson™ AGX Orin Active Heatsink
XHG320 NVIDIA® Jetson™ AGX Orin Passive Heatsink
XHG321 Forge 10G PHY Finned Heatsink
XHG322 Forge 10G PHY Thermal Transfer Plate
XHG323 Thermal plate for NVIDIA® Jetson Orin™ NX (16GB)
XHG324 Passive Finned Heatsink for NVIDIA® Jetson Orin™ NX (16GB)
The XHG301 Passive Heatsink dissipates the heat produced by the module through convection.
• Compatibility: NVIDIA Jetson TX2
• Weight: 50g
• Fin Space: 3 mm
• Thermal Resistance: 0,983 C/W
• Mounting Fasteners: M2.0X0.4, PPHMS, 4 mm long, S.S.
• Mounting Fastener Torque Rating: 1,5 in-ib
• Dimensions: 87 x 50 x 16 mm
The XHG302 Active Heatsink dissipates the heat produced by the module through a fan
• Compatibility: NVIDIA Jetson TX2
• Weight: 59g
• Fan Connector: 4-pin Molex 51021-0400
• Mating Header: 4-pin Molex 53261-0471
• Fin Space: 3 mm
• Thermal Resistance: 1,244 C/W
• Mounting Fasteners: M2.0X0.4, PPHMS, 4mm long, S.S.
• Mounting Fastener Torque Rating: 1,5 in-ib
• Dimensions: 87 x 50 x 16 mm
The XHG303 Passive Heatsink dissipates the heat produced by the module through convection.
• Compatibility: NVIDIA Jetson TX2 4GB / TX2i
• Weight: 50g
• Fin Space: 3 mm
• Thermal Resistance: 0,983 C/W
• Mounting Fasteners: M2.0X0.4, PPHMS, 4 mm long, S.S.
• Mounting Fastener Torque Rating: 1,5 in-ib
• Dimensions: 87 x 50 x 16 mm
The active Heatsink XHG304 dissipates the heat generated by the module via a fan.
• Compatibility: NVIDIA Jetson TX2-4GB / TX2i
• Weight: 59g
• Fan Connector: 4-pin Molex 51021-0400
• Mating Header: 4-pin Molex 53261-0471
• Fin Space: 3 mm
• Thermal Resistance: 1,244 C/W
• Mounting Fasteners: M2.0X0.4, PPHMS, 4 mm long, S.S.
• Mounting Fastener Torque Rating: 1,5 in-ib
• Dimensions: 87 x 50 x 16 mm
Designed to fit the NVIDIA® Jetson AGX Xavier™ modules this passive Heatsink dissipates the heat from the module by means of convection.
• Compatibility: NVIDIA Jetson AGX Xavier
• Thermal Resistance: 0,5 C/W
• Weight: 91g
• Mounting Fasteners: M3.0X0.5 PPHMS, 30mm long, S.S
• Mounting Fastener Torque Rating: 5,2 in-ib
• Dimensions: 100 x 87 x 16 mm
Designed to fit the NVIDIA® Jetson AGX Xavier™ modules this active Heatsink dissipates the heat generated by the module via a fan.
• Compatibility: NVIDIA Jetson AGX Xavier
• Weight: 204g
• Fan Connector: 4-pin Molex 51021-0400
• Mating Header: 4-pin Molex 53261-0471
• Thermal Resistance: 0,5 C/W
• Mounting Fasteners: M3.0X0.5 PPHMS, 30 mm long, S.S
• Mounting Fastener Torque Rating: 5,2 in-ib
• Dimensions: 100 x 87 x 36,7 mm
The Liquid Cooling Block is the only industrial grade water cooling thermal solution for the NVIDIA Jetson platform. Designed to attach directly to NVIDIA Jetson AGX Xavier’s TTP, the Liquid Cooling Block provides thermal relief to the module’s GPU, CPU, and memory components. Liquid Cooling is ideal for applications where sufficient airflow is unavailable, or any high-powered compute system that requires noise reduction or additional cooling. The Liquid Cooling Block can be plumbed into any size of radiator and has the flexibility of adjusting the inlet/outlet port selection.
Engineered for rugged and compact applications, the cooling block features a total of 8 low profile side ports that allow for versatile connectivity options. Optimized for even flow distribution, this Liquid Cooling Block will retain performance regardless of which position the inlet and outlet streams are oriented.
100 x 87 x 18 mm
Dimensions
Input/Output Ports
Included Fittings
Weight
Mounting Hardware
Mounting Fasteners
(20,2 mm to top screws)
8x customizable side ports for low profile usage
• 2x ports per side, G1/8” threaded
2x G1/8” 6 mm barb
6x G1/8” plug
231g
Standoff: M3.0X0.5, 18 mm, QTY4
Spacer: M3.0, 4 mm, QTY4
Spacer: M3.0, 1 mm, QTY4
Screw: M3.0X0.5, PPHMS, 45 mm, QTY4
Thermal Pad: 60 x 60 x 1 mm
M3.0X0.5, PPHMS, length 45 mm
Fastener Torque Rating 5,2 in-lb
Available Integrations
(for details refer to the Connect Tech website under Ordering information)
The Liquid Cooling Block can be integrated with the Rogue or Rogue-X Carrier Board. Optional assemblies come complete with a pre-installed, flashed NVIDIA Jetson AGX module which can be combined with userpreferred add-ons such as additional storage, WiFi, and Bluetooth options.
Industrial grade Liquid Cooling Block for production deployment. Comes equipped with thermal materials and mounting equipment to directly attach to the NVIDIA Jetson AGX Xavier TTP.
The Liquid Cooling Development Kit is intended for laboratory testing only and comes with all materials needed for testing and verification. In addition to the liquid cooling block, the Development Kit also comes with tubing, reservoir, pump, fan, radiator, and coolant concentrate. As these additional components are commercially available, they are not recommended for production environments.
If you are looking for a complete production solution, The Liquid Cooling Block can be bundled with the Rogue or Rogue-X Carrier Board and supplied with the NVIDIA Jetson AGX Xavier module pre-installed and a choice of optional features such as Storage, Bluetooth and WiFi. CTI also flashes the module to enable a fast time-to-market. Contact us for further details.
The Liquid Cooling Development Kit comes supplied with all the materials needed for testing and verifying liquid cooling in a lab environment.
• Liquid Cooling Block
• Reservoir & Pump (160 SPC PWM)
• Fan (500-2200rpm)
Included Components
• Radiator (120 single)
• Rubber Tubing (2m)
• Coolant Concentrate (100mL)
• Clamps, assembly tools, and cables
Designed to fit the NVIDIA® Jetson Nano™ modules this passive Heatsink dissipates the heat generated by the module by means of convection.
• Compatibility: NVIDIA Jetson Nano
• Fin Space: 3mm
• Weight: 34g
• Mountain Fasteners: Alloy Steel M3.0X0.5, Ultra Low Profile 6 mm long
• Mountain Fasteners Torque Rating: 3,1 in-ib
• Dimensions: 61 x 40 x 16 mm
The XHG309 unit was specifically developed for the NVIDIA® Jetson Nano™ modules and dissipates the heat generated by the module via a fan.
• Compatibility: NVIDIA Jetson Nano
• Fin Space: 3 mm
• Weight: 46g
• Mountain Fasteners: M3.0X0.5, Ultra Low Profile 6,0mm long, Alloy Steel
• Mountain Fasteners Torque Rating: 3,1 in-ib
• Dimensions: 61 x 40 x 16 mm* / Note: *Screws of fan protrude past 16 mm
Designed to fit the NVIDIA® Jetson Nano™ production modules NVIDIA® Jetson Nano™ production module does not include a thermal transfer plate With the XHG310, CTI offers a thermal transfer plate solution that users can integrate into a custom enclosure or to other interface material for further cooling.
• Compatibility: NVIDIA Jetson Nano
• Weight:28g
• Mounting Fasteners: M3.0X0.5, Ultra Low Profile 6 mm long, Alloy Steel
• Mounting Fasteners Torque Rating: 3,1 in-ib
• Dimensions: 61 x 40 x 4,2 mm
Designed to fit the NVIDIA® Jetson Xavier™ NX modules The passive Heatsink dissipates the heat generated by the module by convection.
• Compatibility: NVIDIA Jetson Xavier NX
• Weight: 47g
• Mounting Fasteners: Shoulder Screw, M2.0X0.4, 3 mm Dia., 4mm Long Shoulder, Stainless Steel
• Mounting Fasteners Torque Rating: 1,5 in-lb
• Dimensions: 63 x 40 x 16 mm
The NVIDIA® Jetson Xavier™ NX provides supercomputing power in a small footprint. As this module does not have integrated thermal solutions as standard, additional thermal considerations are required for all projects. To get your project off the ground, CTI has developed a selection of out-of-the-box thermal options to get you started in record time. There are two standard active Heatsinks for Xavier NX projects. Both are designed for the NVIDIA® Jetson Xavier NX™ modules; the XHG312 and the XHG314. Both dissipate the heat generated by the module via a fan. The differences between the specifications of the two Heatsink options for Xavier NX projects can be found in the Heatsink comparison table below.
• Compatibility: NVIDIA Jetson Xavier NX
• Aluminium Alloy black finish
• Weight: 57g
• Mounting Fasteners: Shoulder Screw, M2.0X0.4, 3mm Dia., 4 mm Long Shoulder, Stainless Steel + 1 Leaf Spring
• Mounting Fasteners Torque Rating: 1,5 in-lb
• Dimensions: 63 x 40 x 16 mm (Screws of fan protrude past 16 mm) XHG312
for Memory
for Power Components
The XHG312 is a high-end active Heatsink that offers maximum heat dissipation with the smallest possible footprint.
• Compatibility: NVIDIA Jetson Xavier NX
• Aluminium Alloy black finish
• Weight: 57g
• Mounting Fasteners: Shoulder Screw, M2.0X0.4, 3mm Dia., 4 mm Long Shoulder, Stainless Steel + 1 Leaf Spring
• Mounting Fasteners Torque Rating: 1,5 in-lb
• Dimensions: 63 x 40 x 16 mm (Screws of fan protrude past 16 mm)
Cooling options for NVIDIA® Jetson Xavier™ NX projects
The XHG313 is designed specifically for the NVIDIA® Jetson Xavier™ NX production module and can be integrated into a custom enclosure or other surface materials for further cooling.
• Compatibility: NVIDIA Jetson Xavier NX
• Aluminium Alloy black finish
• Weight: 30g
• Mounting Fasteners: Shoulder Screw, M2.0X0.4, 3 mm Dia., 4mm Long Shoulder, Stainless Steel
• Mounting Fasteners Torque Rating: 1,5 in-lb
• Dimensions: 63 x 38,8 x 4,7 mm
Note: NVIDIA® Jetson Xavier™ NX production module does not include an integrated thermal transfer plate out of the box.
The XHG314 is an active Heatsink for projects with non-complex thermal and spatial requirements. It was developed for the NVIDIA® Jetson Xavier NX™ modules. The active Heatsink dissipates the heat generated by the module via a fan.
• Compatibility: NVIDIA Jetson Xavier NX
• Weight: 50g
• Mounting Fasteners: Shoulder Screw, M2.0X0.4, 3 mm Dia., 4 mm Long Shoulder, Stainless Steel + 1 Leaf Spring
• Mounting Fasteners Torque Rating: 1,5 in-lb
• Dimensions: 57,8 x 39 x 18 mm
Note: NVIDIA® Jetson Xavier™ NX production module does not include an integrated thermal transfer plate out of the box.
Designed to fit the NVIDIA® Jetson™ TX2 NX module and protect the TX2 NX GPU die, the thermal transfer plate conduction-cools the heat produced by the module.
A customer would order this plate if designing their own thermal solution where they mount the module on the wall of an enclosure, for example, using a much larger surface area to dissipate heat. It cannot be used with an active or passive Heatsink.
• Compatibility: NVIDIA Jetson TX2 NX
• Aluminium Alloy black finish
• Weight: 35g (Heatsink only)
• Mounting Fasteners: (4X) Shoulder Screw M2.0X0.4, Dia. 3 mm, long 3 mm + (1X) leaf spring; mass/weight 3g
• Dimensions: 64 x 39 x 8 mm
Note: NVIDIA® Jetson Xavier™ NX production module does not include an integrated thermal transfer plate out of the box.
Designed to fit the NVIDIA® Jetson™ TX2 NX module and protect the TX2 NX GPU die. It dissipates the heat generated by the module through convection.
• Compatibility: NVIDIA Jetson TX2 NX
• Aluminium Alloy black finish
• Weight: 44g (Heatsink only)
• Mounting Fasteners: (4X) Shoulder Screw M2.0X0.4, Dia. 3 mm, long 3 mm + (1X) leaf spring; mass/weight 3g
• Dimensions: 63 x 40 x 18 mm
The active Heatsink was specially developed for the NVIDIA® Jetson™ TX2 NX module and protects the TX2 NX GPU die. It dissipates the heat generated by the module by means of a fan.
• Compatibility: NVIDIA Jetson TX2 NX5
• Mounting Fasteners: Leaf spring, 4 x M2 screws
• Dimensions: 30 x 35 x 13,5 mm
XHG319 XHG320 XHG321 XHG322
XHG319 AGX Orin Active Thermal
XHG320 AGX Orin Active Thermal
XHG321 Forge 10G PHY Finned Heatsink
XHG322 Forge 10G PHY Thermal Transfer Plate (TTP). Note: This is pre-installed on all AGX201 / AGX201-XX builds
Connect Tech recommends attaching both the 10G PHY TTP and the NVIDIA Jetson Module TTP to a surface that can provide adequate relief to keep the top surfaces of both TTPs below 80 C for conditioning cooling.
The NVIDIA® Jetson Orin™ NX production module does not include a heat transfer plate. Connect Tech offers a thermal transfer plate that users may attach into a bespoke chassis or other interface material to help cool the Jetson Orin™ NX manufacturing module. Connect Tech offers a selection of heatsinks, allowing you to be up and running quickly.
• Mounting Fasteners: M3.0X0.5, Ultra Low Profile 6,0 mm long, Alloy Steel
• Mounting Fasteners Torque Rating: 3.1 in-ib
• Designed to fit the NVIDIA® Jetson Orin™ NX production modules
• NVIDIA® Jetson Orin™ NX production module does not include a thermal transfer plate
• CTI provides Thermal transfer plate that users can integrate into a custom chassis or to other interface material for further cooling
• Dimensions: 63mm x 39mm x 4,7mm
XHG324 Passive Finned Heatsink for NVIDIA® Jetson Orin NX (16GB) / Orin Nano
The NVIDIA® Jetson Orin™ Nano is a compact, potent computer for embedded applications and AI IoT that provides the power of contemporary AI. Connect Tech offers embedded system solutions and compact form factor carriers for the Jetson Orin™ Nano. Accessories that complement their products are available to aid in product selection. Quickly get operational with readily available thermal solutions.
The XHG324 Heat Sink, which is compatible with Jetson Orin™ NX modules, disperses heat via convection.
• Designed to fit NVIDIA® Jetson Orin™ NX modules
• Dissipates the heat produced by the module through convection
• Dimensions: 63 x 40 x 16 mm
XHG325 Active Thermal Heatsink for NVIDIA® Jetson Orin NX (16GB) / Orin Nano
NVIDIA® Jetson Orin™ NX delivers supercomputer performance in a small package. Because this module does not come with any integrated thermal solutions, all projects will necessitate additional thermal consideration. In order to get your project off the ground as quickly as possible, Connect Tech has designed a variety of easily available thermal alternatives.
• Designed to fit the NVIDIA® Jetson Orin NX™ modules
• Two standard Active Heat Sinks to choose from
• Dissipates the heat produced by the module through a fan
• Mounting Fasteners Torque Rating: 1.5 in-lb
EMBEDDED SYSTEMS – NVIDIA JETSON AGX ORIN™
CORE01/CORE02 Essential EdgeAI - AGX Orin™
ESG620/ESG621 Anvil - Jetson AGX Orin™
ESG630 Sentry-X2 Rugged System - AGX Orin™ Industrial
ESG635 Anvil-RX – AGX Orin™
UAGX2U Inference Server - AGX Orin™
EMBEDDED SYSTEMS – NVIDIA JETSON ORIN™ NX & ORIN™ NANO
CORE03/CORE04 Essential EdgeAI - Orin™ NX & Orin™ Nano
ESG604/ESG608 Polaris – Jetson Orin™ NX
ESG605 Rudi-NX – Jetson Xavier™ NX
ESG606 Rudi-NX FPD-Link III – Orin™ NX
UAGX2U Inference Server - AGX Orin™
EMBEDDED SYSTEMS – NVIDIA JETSON XAVIER™
ESG602 Rudi-NX
ESG603 Rudi-NX FPD-Link III – Xavier™ NX
ESG610 Rudi-AGX - AGX Xavier™
SGX001/SGX011 Sentry-X Rugged System – AGX Xavier™
VPG004 Graphite VPX
PART NUMBER OVERVIEW
PRODUCT CHANGE NOTIFICATIONS (PCN)
NVIDIA JETSON DEVELOPER KITS & MODULES
CORE01-02
CORE01-06
CORE02-02
CORE02-06
CORE03-02
CORE03-05
CORE03-08
CORE03-11
CORE04-02
CORE04-05
CORE04-08
CORE04-11
CVS001
D7G001
D7G002
D7G003
D7G004
D7G005
D7G006
DXG201
DXG202
DXG203
DXG204
DXG301
DXG302
ESG602
ESG603
ESG604
ESG605
ESG606
ESG610
ESG620
ESG621
ESG630
ESG635
ESG701
ESG702
ESG703
ESG704
ESG705
ESG706
ESG710
ESG711
ESG712
SGX001
SGX001-05
SGX001-06
SGX011
UAGX2U
V7G001
V7G002
V7G003
V7G004
V7G005
V7G006
VPG004-01
VPG004-02
VXG201
VXG202
VXG203
VXG204
VXG301
VXG302
Over the last year, Connect Tech launched two embedded systems: Anvil and Polaris. Though both are strong and feature-rich systems in their own right, they differ greatly in ruggedness, performance, and I/O. Designed for the strong NVIDIA Jetson AGX Orin and Jetson AGX Orin Industrial modules, Anvil; Polaris was built for the sturdy Jetson Orin NX and Orin Nano. Rated IP67, Polaris's companion is ready for its own tough version. While Polaris offers 16GB, Anvil offers up to 64GB. While Anvil lets for 8x GMSL camera in, Polaris offers 4x GMSL connections.
As an introduction to both Embedded Systems, Connect Tech has prepared two informative video presentations (To view these, click on the camera symbols below):
What industries are Anvil and Polaris best suited for?
Understandably, Anvil and Polaris are both suited for a wide range of applications across numerous industries. See the list of targeted applications below.
OMPARISON CHART
Essential EdgeAI solutions are designed for generative AI and AI-enabled video analytics at scale. Each system supports active or passive thermal management and incorporates USB 3.2, Gigabit Ethernet, an NVMe M.2 M-Key 2280 storage device, and a wide input power range. While the Essential EdgeAI network nodes are a low-cost option, they additionally include the Connect Tech Board Support Package (BSP) support, which ensures that large-scale installations go smoothly from start to finish.
• Get your pick of the NVIDIA® Jetson Orin™ Nano, Jetson Orin™ NX, or Jetson AGX Orin™ preintegrated solutions.
• Choice of Active and Passive Thermal Cooling
• Includes key interfaces: USB 3.2, GbE, and UART
• Scalable AI solution
• Small footprint
Not the perfect fit? Engage with Jetson experts to guarantee successful large deployment.
CORE - Essential EdgeAI integrated with NVIDIA® Jetson AGX Orin™
CORE - Essential EdgeAI Recommended cables/accessories
Essential EdgeAI solutions are designed for generative AI and AI-enabled video analytics at scale. Each system supports active or passive thermal management and incorporates USB 3.2, Gigabit Ethernet, an NVMe M.2 M-Key 2280 storage device, and a wide input power range. While the Essential EdgeAI network nodes are a low-cost option, they additionally include the Connect Tech Board Support Package (BSP) support, which ensures that large-scale installations go smoothly from start to finish.
• Get your pick of the NVIDIA® Jetson Orin™ Nano, Jetson Orin™ NX, or Jetson AGX Orin™ pre-integrated solutions.
• Choice of Active and Passive Thermal Cooling
• Includes key interfaces: USB 3.2, GbE, and UART
• Scalable AI solution
• Small footprint
Cables and accessories
CBG788 Serial Debug Cable, USB-UART 3.3V Cable MSG120 Power Supply, Jetson Orin NX/Nano variant
Not the perfect fit? Engage with Jetson experts to guarantee successful large deployment.
• Extremely small footprint: 135 x 50 x 105 mm
• GMSL, USB 3.0, USB 2.0, CAN 2.0b, USB OTG, RS-485, I2C, GPIO, SPI, PWM
• 1x NVMe (PCIe x4, 2280), 1x SD card slot
• Operating Temperature Range: -20 to +80 C
Connect Tech’s Rudi-NX Embedded System utilizes the new NVIDIA® Jetson Xavier™ NX, the World’s Smallest AI Supercomputer. Delivering up to 21 TOPS of performance in 15W. With 384 CUDA Cores, 48 Tensor Cores, and 2 NVDLA engines, it can run multiple modern neural networks in parallel and process high-resolution data from multiple sensors simultaneously.
The Rudi-NX enables multiple vision inputs including up to 4x GMSL (MIPI CSI-2) Camera inputs, 4x USB 3.0 ports and 2x Gigabit Ethernet ports. Additional connectivity via CAN, UARTs, RS-485, I2C, SPI, and GPIO.
Important Note: The ESG602 series has no active fans installed. All ESG602 Rudi Embedded units that come pre-installed with NVMe have a passive thermal plate and a different base plate other than the standard ESG602-01. If NVMe is added later to an ESG602-01, then the customer is responsible to finding their own thermal solution.
Rudi-NX GMSL Embedded System
Part No.
Rudi-NX GMSL System Jetson Xavier NX 16GB
ESG602-01 X X ESG602-04 X X 1TB X ESG602-07 X X 1TB X
ESG602-08 X X 1TB X X
ESG602-19 X X 1TB
The above integrated solutions are for the EMEA region, for other regions, please contact us. Not the perfect fit? Contact us to engage with Connect Tech’s Jetson experts to guarantee successful large-scale deployment.
Rudi-NX GMSL Embedded System Accessories
Part No. Description
CBG341 Quad FAKRA GMSL Cable, 1 metre length
MSG067 Rudi Embedded System External Mounting Brackets (2 Brackets included). No additional mounting HW is required
MSG085 Power Supply 19V 120W AC/DC External Desktop PSU (Class I) with CBG315 Input Power Cable and CBG411 4-pin DIN to 4-Pin Mini-Fit Jr. suitable for Rogue, Rudi-NX and Rudi-AGX.
MSG093 WiFi/BT Antenna
MSG094 LTE Antenna
MSG095 GPS Active External Antenna with 3m RG174 Cable + SMA Male Connector
PAR502 WiFi Modules (802.11) Intel Dual Band Wireless-AC 7265
• Extremely small footprint: 135mm x 50mm x 105mm
• 4x FPD Link, USB 3.0, USB 2.0, CAN 2.0b, USB OTG, RS-485, I2C, GPIO, SPI, PWM
• 1x NVMe (PCIe x4, 2280), 1x SD card slot
• Operating Temperature Range: -20 to +80 C
The Rudi-NX from Connect Tech is the ultimate Edge AI computing device for cutting-edge, compute-intensive applications. It is capable of delivering up to 21 TOPS of performance in 15W. It can run many current neural networks in parallel and analyse high-resolution data from multiple sensors at the same time thanks to its 384 CUDA Cores, 48 Tensor Cores, and 2 NVDLA engines.
The Rudi-NX supports up to four GMSL (MIPI CSI-2) camera inputs, four USB 3.0 ports, and two Gigabit Ethernet interfaces. CAN, UARTs, RS-485, I2C, SPI, and GPIO provide further connectivity.
• IP67 Rated, Actively Cooled
• Rugged M12: 2x GbE, 2x CAN, GPIO
• Wide range isolated power input (+18V – +48V)
• 2x USB3.1
• 4x GMSL2 via sealed FAKRA
• 4G/5G/LTE, WiFi/BT, GNSS, M.2 2280 NVME M-Key
• Operating Temp.: -20 to +80 C
• Dimensions: 285 x 150 x 70 mm
The robust Polaris system, powered by the Jetson Orin™ NX and designed for robotics, smart cities, and autonomous machines, offers a wide range of connectivity in an IP67-rated rugged chassis.
ESG604 - Polaris Embedded System with NVIDIA® Jetson Orin™ NX module Part No.
Jetson module Thermal Cooling options
Passive Active
ESG604-1C Orin NX 16GB X 2TB
ESG604-1D Orin NX 16GB X 2TB WiFi
ESG604 - Polaris Embedded System - Recommended Optional Accessories Part No.
Recommended Antennas
LTE/WiFi/GNSS MA850.A.LBICG.002 TAOGLAS 5G/LTE/GNSS YB0007AA Quetel
LTE TG.66.0723 Taoglas
WiFi GW.05.AE23 Taoglas
GNSS AA.105.301621 Taoglas
• 4x GMSL, USB 3.0, USB 2.0, CAN 2.0b, USB OTG, RS-485, I2C, GPIO, SPI, PWM
• 1x NVMe Expansion (M.2 key - PCIe x4, 2280)
• Operating Temperature Range: -20 to +80 C
• Extremely small footprint: 135mm x 50mm x 105mm
The Rudi-NX from Connect Tech is the pinnacle of Edge AI computing for cutting-edge, computationally demanding applications. It can give up to 100 TOPS of performance in 15W when powered by NVIDIA® Jetson Orin™ NX. It can process high-resolution data from numerous sensors at once and run several contemporary neural networks in parallel thanks to its 1024 CUDA Cores, 32 Tensor Cores, and two NVDLA v2 engines.
Up to four GMSL (MIPI CSI-2) camera inputs, four USB 3.0 ports, and two Gigabit Ethernet ports are all supported by the Rudi-NX. Further communication through RS-485, I2C, SPI, GPIO, CAN, UARTs, and CAN.
Note: The User Manual for this product has not been released yet. Please contact us for further information and to request the link when it has been published.
Caution: The Jetson Orin NX Block Diagram, published by NVIDIA, illustrates what connectors the NX Orin offer. Since these may differ from those offered by Connect Tech’s Carrier Board, please consult the User Manual for exact details.
The Rudi NX for Orin NX contains thermocouples designed specifically for the Rudi. The housing acts as a thermal, and there is also an internal heat transfer plate that transfers heat from the internal components to the Rudi housing.
These are different from our board level active and passive thermostats that you can find on Connect Tech’s website.
ESG605 - Rudi-AGX Embedded System with NVIDIA® Jetson AGX Xavier™ module
ESG605 - Rudi-AGX Embedded System - Recommended Optional Accessories Part No.
CBG341 Quad FAKRA GMSL1/2 Cable
CBG408 Rudi-AGX Power Connector Matin Cable
CBG629 GPIO Cable - 40-pin Tiger-Eye to Unterminated Flying Leads - SFSD-20-28C-G-12.00-SR
CBG814 CAN Cable - SFSD-03-28C-G-12.00-SR Diverse
MSG067 Mounting Brackets
MSG093 WiFi/BT Antenna
MSG094 LTE Antenna Power Adapter
MSG085 Rudi-AGX AC/DC Power Supply
• 4x FPD Link, USB 3.0, USB 2.0, CAN 2.0b, USB OTG, RS-485, I2C, GPIO, SPI, PWM
• 1x NVMe Expansion (M.2 key - PCIe x4, 2280)
• Operating Temperature Range: -20 to +80 C
• Extremely small footprint: 135mm x 50mm x 105mm
For cutting-edge, computationally demanding applications, Connect Tech's Rudi-NX represents the apex of edge AI computing. When powered by NVIDIA® Jetson Orin™ NX, it can deliver up to 100 TOPS of performance in 15W. Its 1024 CUDA Cores, 32 Tensor Cores, and two NVDLA v2 engines enable it to handle high-resolution data from several sensors simultaneously and run several modern neural networks in parallel.
The Rudi-NX supports up to four GMSL (MIPI CSI-2) camera inputs, four USB 3.0 ports, and two Gigabit Ethernet interfaces. Additional communication through CAN, UARTs, RS-485, I2C, SPI, GPIO, and SPI.
Note: The User Manual for this product has not been released yet. Please contact us for further information and to request the link when it has been published.
The Rudi AGX Embedded System is a full-featured and thermally optimized, fan-less embedded system can instantly deploy any advanced AI application.
As NVIDIA® Jetson’s largest hardware partner Connect Tech is constantly striving to develop off-the-shelf solutions that support the Jetson community’s production goals. As the most powerful SoC in the Jetson line-up, the AGX Xavier empowers customers to deploy state-of-theart and computationally heavy AI workloads at the Edge. Today we’re unveiling our latest commercially available edge device: Rudi-AGX.
Rudi-AGX unleashes the full potential of the Jetson AGX Xavier module. It is capable of running AI programs at Maximum Performance (MAXN) while maintaining a safe operating temperature. Rudi-AGX is the edge device designed to seamlessly deploy the next generation of autonomous vehicles, smart city applications, and intelligent video solutions.
With a AGX Xavier SoM embedded, Rudi-AGX features a Volta GPU with 512 CUDA® cores, 64 Tensor Cores, and capable of achieving an AI performance of 32 TOPs. Not all AGX Xavier systems are built the same. Rudi-AGX‘s thermal optimization allows access to the module’s full AI performance capabilities, making it the AI Edge solution compute-intensive applications rely on.
Designed to accommodate sensor heavy applications, Rudi-AGX provides the ability to directly connect eight GMSL cameras, expand storage with NVMe or an SD card, and utilize a video capture card. Connectivity will not be a problem with WiFi, Bluetooth, and LTE slots, as well as two GbE ports available. This feature rich edge device also comes equipped with 6x GPIO, 4x USB 3.0, 2x UART, 2x I2C, 2x SPI, and an RS-485 connector. Rudi-AGX is ready for instant field deployment with any program developed using NVIDIA’s JetPack SDK or deep learning toolkit application.
Don’t sacrifice AI performance due to thermal limitations. Rudi-AGX will not throttle performance within the operating temperature range when running in 30W mode with 0 CFM airflow, or with 125 CFM airflow at MAX-N.
ESG610 - Rudi-AGX Embedded System with NVIDIA® Jetson AGX Xavier™ module
ESG610 - Rudi-AGX Embedded System - Recommended Optional Accessories
Rudi-AGX Orin unleashes the full potential of the NVIDIA® Jetson AGX Orin™ module, an Industrial Embedded AI system ideal for a range of Edge Computing applications.
Features:
• Compatibility: NVIDIA® Jetson AGX Orin™ 32Gb or 64GB Module
• Option for either Active (ESG620) or Passive (ESG621) cooling
• Networking: 2x 10GBASE-T (10G Ethernet) | 1000BASE-T (1G Ethernet)
• 1x DisplayPort
• Camera Inputs: 1x 16-Lane MIPI Expansion Connector
• Interface directly to a range of MIPI CSI, GMSL2 r FPD-Link III image sensors.
• USB: 1x USB 3.2 Ports (Type-C – OTG Capable), 1x USB 3.2 Ports (Type-C)
• Storage: 1x Micro SD, 2x NVMe M.2 Key M Slots
• UART: 2x @ 3.3V levels UART1 and UART2 | 2x RS-232/485 UARTs
• I2C/SPI: 2x I2C Channel @ 3.3V IO | 2x SPI Channel @ 3.3V IO
• CAN: 2x CAN 2.0b Non-Isolated Port
• Input Power: +10 to 36V DC Wide Input Power (6-pin Mini-fit Jr. Connector)
• GPIO: 4x GPI, 4x GPO User Expansion: 1x M.2 Key E 2230 Wifi/BT (PCIe/USB2) 1x M2 Key B 3042/3052 (LTE/5G – USB3) with Micro SIM
• Buttons: Power, Reset, Recovery Tactile Buttons as well as fine pitch pin header connector)
• RTC Battery: 3-pin Picoblade connector to accept CTI battery cable assembly
• Dimensions: 155 x 125mm
• Operating Temperature: -40 to +85C
With its power-efficient and feature-rich design, Anvil is ready to handle even the most compute-intensive AI applications. With this cutting-edge NVIDIA® Jetson AGX Orin™ supercomputer, you can easily deploy your next generation autonomous vehicle, smart city application, or intelligent vision solution. Custom integrations with external storage, LTE, 5G, and WiFi/Bluetooth are also available via Intel 8265.
Anvil System Embedded System Configuration options
ESG620/621 - Anvil Embedded System for AGX Orin - Recommended cables/accessories Part No. Description
Cabling
CBG136 CR2032 RTC Battery w/ 3-pin Connector Cable Assembly - 6032101013, 70mm
CBG341 Quad FAKRA GMSL1/2 Cable
MSG067 Mounting Brackets
MSG093 WiFi/BT Antenna
Diverse
MSG094 LTE Antenna Power Adapter
MSG103 PSU 6pin 12V 250W (CTI sourced power brick)
• 2x 3G-SDI/HD-SDI/SDI Inputs
• 2x 3G-SDI/HD-SDI/SDI Outputs
• Up to 4TB NVMe Storage
• 3x Ethernet Ports: 10GbE, 2.5GbE, 1GbE
• Tested to IP67, MIL-STD-810H
ESG630 - Sentry-X2 Embedded System integated with NVIDIA® Jetson AGX Orin™ Industrial Module Part No. Jetson module Thermal Cooling options
ESG630-02 AGX Orin 64GB Ind. X 2TB 4x 3G-SDI inputs
ESG630-11 AGX Orin 64GB Ind. X 2TB 4x 3G-SDI inputs
Powered by the NVIDIA® Jetson AGX OrinTM Industrial module, Sentry-X2 is a MIL-Rugged system. ESG630 Sentry-X2 is best suited for avionic applications.
The Sentry-X2 is a MIL-SPEC certified artificial intelligence system for tactical edge computing, built for mission-critical applications in challenging situations. A 2048-core GPU with 248 TOPS, the NVIDIA® Jetson AGX OrinTM Industrial is perfect for sophisticated AI and machine learning applications. The Sentry-X2 has two 3G-SDI/HD-SDI/SDI inputs and two 3G-SDI/HD-SDI/SDI outputs, making it appropriate for complicated vision systems. It also has two USB 3.2 ports, 10Gb, 2.5Gb, and 1Gb Ethernet, and a variety of I/O choices.
Sentry-X2 is made to satisfy the exacting standards of heavy industrial, military, and aerospace applications. It is tested to MIL-STD criteria for stress and vibration and uses IP67 certified D38999 connections for data integrity.
Anvil-RX elevates tough AI and powerful computing performance to new heights.
Connect Tech's Anvil-RX is a rugged embedded system that uses the NVIDIA® Jetson AGX Orin™. This powerful edge system is fully IP67 protected and comes with M12 and sealed PC ports, making it ideal for mission-critical applications in hostile locations. The Anvil-RX is best suited land-based vehicles, Autonomous Mobile Robot (AMR), Automated Guided Vehicles (AGVs) in particular in agriculture, and industrial applications. For MIL STD please contact us.
• IP67 Rated
• Rugged M12 Connectors
• 8x GMSL2 Camera Inputs
• Up to 2x NVMe (2/4/8TB)
• 2x USB 3.2, 2x CAN, HDMI, UART, GNSS
• Storage: 64GB eMMC (on Jetson Modules) and up to 2 x NVMe (2/4/8TB)
• Dimensions: 279 x 214 x 145mm
• Operating Temp.: -40 to +65 C
ESG635 - Anvil-RX Embedded System integrated with NVIDIA® Jetson AGX Orin™
ESG635 - Embedded Systems - Recommended cables/accessories
Cabling
CBG756 Anvil-RX CAN Cable (M12 5-Pin)
CBG757 Breakout/Peripherals End - Connector Type Flying Lead Anvil-RX I/O Cable: I2C, SPI (M12 8-Pin)
CBG758 Anvil-RX I/O Cable: GPIO (M12 12-Pin)
CBG759 Anvil-RX Flying Lead (M12 5-Pin) Diverse
MSG093 WiFi/BT Antenna
MSG094 LTE Antenna
MSG114 Anvil-RX Dust Cap Kit Power Adapter
MSG116 Anvil-RX AC/DC Power Supply
• 12x 275 TOPS, 2048-core NVIDIA® Ampere™ GPU and 64 Tensor Cores
• 4x 10G SFP+ uplink capability
• Supports External NVMe Storage
• 2U ATX style redundant power supply
UAGX2U - NVIDIA® Jetson AGX 2U Inference Server with 12x AGX Orin™ 32GB Modules
High-powered computing gains flexibility and cost economy from GPU -Accelerated Machine Learning and Inferencing Servers.
Designed in association with USES Integrated Solutions, the AGX Orin Inference Server is a very low wattage, high performance AI workstation driven by the NVIDIA Jetson AGX Orin 64GB module. Running NVIDIA's most potent deep -learning software libraries, this inference server addresses the difficulty of widely distributing edge solutions. Via a dedicated Managed Ethernet Switch with 10G uplink capacity, a Gigabit Ethernet fabric link s all of the on-board 12x Jetson AGX Orin modules.
SGX001 / SGX01
• Unique docking stations – allows for quick release of Sentry-X compute module
• Ideal for Aerospace and Defence applications
• Tested to MIL-STD-810G, DO-160G for shock & vibration, ingress protection of IP67
• Wide I/O assortment including 2x GbE, 3x USB 3.1, 2x HDMI, 2x CAN 2.0b, 4x GPI, 4x GPO, 2x RS-232/422/485
Connect Tech's Sentry-X Rugged Embedded System is designed for the NVIDIA® Jetson AGX Xavier™, providing a diverse range of I/O. This embedded system comes with 2x GbE, 3x USB 3.1, 2x HDMI, 2x CAN 2.0b, 1x RS-232, 1x RS-422/485, 1x UART 3.3V TTL, and 8x GPIO.
Sentry-X has been tested to MIL-STD-810G and DO-160G stress and vibration standards, as well as IP67 ingress protection.
The embedded system is suited for Aerospace and Defence applications, as well as any market that may benefit from the exceptional performance of the Jetson AGX Xavier in a durable enclosure.
Sentry-X Rugged Embedded System powered by NVIDIA® Jetson AGX Xavier™
Key Power Cable (2.5mm barrel socket
CBG336 Fischer MiniMax Series 08 Shell 24 Pin 02 Key Serial/GPIO Cable (flying leads)
CBG337 Fischer MiniMax Series 08 Shell 24 Pin 01 Key HDMI Cable (HDMI Type A plug)
CBG339 Fischer MiniMax Series 08 Shell 08 Pin 02 Key CAN Bus Cable (flying leads)
CBG523 Meritec Hercules 38999 size 09 Plug, Cadmium plated, “N” key shell to Pigtails (flying leads)
CBG524 Meritec Hercules 38999 size 17 Plug, Cadmium plated, “N” key shell to Pigtails (flying leads)
CBG525 Meritec Hercules DisplayPort Connector
CBG526 Meritec Hercules 38999 size 23 Plug, Cadmium plated, “N” key shell to (2) HDMI (Socket) cable, (2) USB 3.0 Type A (Socket) cable, and (2) RJ45 (Socket) UTP CAT6 cable Power Adapter
MSG089 AC/DC Power Supply for Sentry-X Rugged Embedded System. 100-240VAC Input, +24V DC Output, 220W
ESG701 ESG702 ESG703 ESG704 ESG705 ESG706 ESG710 ESG711 ESG712
Connect Tech's V7G GPU System combines Intel® Xeon® D (Server Class) and Intel® Atom™ C3000 x86 processors with high-end NVIDIA® Quadro® and Tesla® graphics processing units (GPUs) in a small form factor embedded system. You have a choice of the highest performance and lower performance models, all ideal for high-end video encoding/decoding applications or GPGPU CUDA® processing, deep learning and artificial intelligence applications.
This embedded system has all the latest generation interconnects including: 10GbE and Gigabit Ethernet, USB 3.0 and 2.0, HDMI, SATA III, GPIO, I2C, M.2, miniPCIe. The black aluminium housing offers two mounting options: Half-rack rail mounting or standalone mounting bracket.
• High-End GPUs paired with Intel® Xeon® D Server Class and Intel® Atom™ C3000 x86 Processors
• 4 independent display outputs or GPGPU processing system using CUDA® cores
• Designed with NVIDIA® Jetson AGX Orin™ Industrial
• 3U VPX SOSA aligned
• PCIe x8 Gen4 (Endpoint or Host)
• 2x 10GBASE-KR Ethernet
• Complete IPMI Support
The Graphite VPX is a durable 3U VPX single board computer engineered using the NVIDIA® Jetson AGX Orin . Conforming to the Sensor Open Systems Architecture (SOSA) standard, it guarantees smooth interoperability with SOSA-compliant systems.
The Graphite VPX, featuring comprehensive IPMI compatibility, facilitates effective remote management and monitoring function alities. High-speed connectivity options, including twin 10GBASE-KR Ethernet, dual USB 3.2 Gen2, and PCIe x8 Gen4, facilitate substantial expansion and integration. Complying with MIL-STD-810H and DO-160 for stress and vibration, it is an optimal selection for contemporary mission -critical aerospace and defence AI applications.
P/N
Description
VPG004-01 Graphite VPX – Powered by the Jetson AGX Orin™ Module (PCIe Host Mode) Two x4 downstream ports
VPG004-02 Graphite VPX – Powered by the Jetson AGX Orin™ Module (PCIe Endpoint Mode)
RTG005 RTM for Graphite VPX VPG004
900-13448-0020-000 NVIDIA Jetson Nano Module 4GB
900-13636-0010-000 NVIDIA Jetson TX2 NX Module 4GB
900-13701-0040-000 NVIDIA Jetson AGX Orin 32GB
900-13701-0050-000 NVIDIA Jetson AGX Orin 64GB
900-13701-0080-000 NVIDIA Jetson AGX Orin Industrial Module 64GB
900-13767-0000-000 NVIDIA Jetson Orin NX 16GB
900-13767-0010-000 NVIDIA Jetson Orin NX 8GB
900-13767-0030-000 NVIDIA Jetson Orin Nano 8GB
900-82888-0040-000 NVIDIA Jetson AGX Xavier Module 32GB
900-82888-0050-000 NVIDIA Jetson AGX Xavier Module 64GB
900-82888-0080-000 NVIDIA Jetson AGX Xavier Industrial Module 32GB
900-83310-0001-000 NVIDIA Jetson TX2 Module 8GB
900-83489-0000-000 NVIDIA Jetson TX2i Industrial Module 8GB
900-83489-0080-000 NVIDIA Jetson TX2 Module 4GB
900-83668-0000-000 NVIDIA Jetson Xavier NX Module 8GB
900-83668-0030-000 NVIDIA Jetson Xavier NX Module 16GB
945-13730-0055-000 NVIDIA Jetson AGX Orin Development Kit 64KB
945-13766-0005-000 NVIDIA Jetson Orin Nano Super Developer Kit 4GB
945-14070-0085-000 NVIDIA Jetson Thor Developer Kit
NVIDIA® JETSON™ MODULE COMPARISON: https://connecttech.com/jetson/jetson-module-comparison/ NVIDIA® JETSON LIFECYCLE: https://developer.nvidia.com/embedded/lifecycle
As a rule, NVIDIA issues a Last Time Buy approx. 3 months before one of their modules is announced EOL. However, to be certain you should always check their website.
Connect Tech provides complete support for all Jetson solutions. Their Jetson Gurus can assist you with getting your application up and running. Any technical questions you may have can be addressed by phone or email. Some frequently asked questions can be found in their FAQ section.
The table lists the modules that are currently supported.
https://connecttech.com/jetson/nvidia-jetson-support/
NVIDIA© Jetson™ systems deliver enhanced performance and energy efficiency for operating autonomous machine software, achieving superior speed while consuming less power. Each constitutes a comprehensive System-on-Module (SOM), incorporating a CPU, GPU, PMIC, DRAM, and flash storage - thereby conserving development time and financial resources. Additionally, Jetson modules offer great flexibility - simply choose the appropriate SOM for the application and construct a tailored system to fulfil specific requirements.
High-efficiency, low-energy NVIDIA Jetson modules provide real-time artificial intelligence (AI) performance precisely where it is most required at the Edge. Complex data processing can now occur on small, power-efficient devices, ensuring rapid and precise inference in network-constrained settings.
NVIDIA Jetson developer kits and modules, along with the NVIDIA JetPack SDK, furnish all necessary components for the development and deployment of applications across various use cases utilising a singular software architecture. This significantly lowers software development expenses and provides a scalable AI strategy for autonomous devices, unleashing substantial potential across several sectors, including healthcare, manufacturing, agriculture, and more.
Industrial Modules
Jetson Orin NX 8GB
• 1024-core NVIDIA Ampere GPU
• 6-core Arm® Cortex®
• 8 GB LPDDR5
• Supports NVMe
Jetson AGX Orin 32GB
• 512-Core NVIDIA Volta GPU with Tensor Cores
• (2x) NVDLA Engines
• 8-Core ARM® v8.2 64-Bit Carmel CPU
• 32GB 256-Bit LPDDR4x|136.5GB/s
• 32 GB eMMC 5.1 Flash Storage
• 7-Way VLIW Vision Accelerator Processor
• 32GB eMMC 5.1 Click here to learn more Click here to learn more
From automating company processes and obtaining insight through data analysis to engaging with consumers and workers, AI is generating competitive advantage.
The industry-leading platform for embedded applications and autonomous machines is NVIDIA Jetson™. Included in this ecosystem are sensors, software development kits (SDKs), services, and devices that can hasten development; the Jetson modules, which are tiny, powerful computers; and the NVIDIA JetPack™ SDK, which can accelerate applications. To help businesses create software-defined autonomous machines, Jetson offers the performance and power efficiency they require while being compatible with cloud-native workflows and other NVIDIA platforms.
NVIDIA Jetson Module Comparison: https://connecttech.com/jetson/jetson-module-comparison/
• 128-core NVIDIA Maxwell GPU
• Quad-core ARM A57 CPU
• 4 GB 64-bit LPDDR4
• 10/100/1000BASE-T Ethernet
DEPLOYING AI ACROSS NUMEROUS NEW EDGE SYSTEMS
The NVIDIA Jetson Nano module is creating remarkable new opportunities for edge computing. It provides up to 472 GFLOPS of accelerated computing, capable of executing numerous contemporary neural networks concurrently, and offers the performance necessary to interpret input from various high-resolution sensors—an essential prerequisite for comprehensive AI systems. It is production-ready and compatible with all major AI frameworks. Jetson Nano is the optimal platform for the development of mass-market AI products, including AIoT gateways, smart network video recorders and cameras, consumer robots, and optical inspection devices.
The system-on-module is equipped with the NVIDIA Maxwell GPU and 4 GB of memory, facilitating real-time processing of high-resolution inputs. It provides a distinctive amalgamation of performance, power superiority, and an extensive array of I/O interfaces, ranging from high-speed CSI and PCIe to low-speed I2Cs and GPIOs. Additionally, it accommodates several unique sensor arrays to provide a multitude of applications with remarkable power efficiency, utilising as low as 5 W.
The Jetson Nano is compatible with NVIDIA JetPack, which comprises a board support package (BSP), Linux operating system, NVIDIA CUDA, cuDNN, and TensorRT software libraries for deep learning, computer vision, GPU computing, multimedia processing, and more applications.
The extensive software stack facilitates rapid AI deployment on autonomous devices, decreases complexity, and accelerates time to market. The identical JetPack SDK is utilised throughout the complete NVIDIA Jetson™ product line and is entirely compatible with NVIDIA’s preeminent AI platform for the training and deployment of AI software.
• NVIDIA Pascal architecture with 256 NVIDIA CUDA cores
• Dual-core NVIDIA Denver 2 64-bit CPU and quad-core Arm Cortex-A57 MPCore processor complex
• 4GB 128-bit LPDDR4
• 16 GB eMMC 5.1
• 10/100/1000 Base-T Ethernet
• Voltage Input: 5V
• Module Power: 7.5W | 15W
For low level embedded and edge devices, the NVIDIA® Jetson™ TX2 NX offers the next stage in artificial intelligence capabilities. It offers up to 2.5X the performance of Jetson Nano and shares form-factor and pin compatibility with Jetson Nano and Jetson Xavier™ NX.
Perfect for developing mass-market AI products in the sectors of manufacturing, logistics, retail, service, agriculture, smart city, and healthcare and life sciences, the small, power-efficient Jetson TX2 NX system-on-module (SOM). While using cloud-native technologies to create, deploy, and control the same software across all Jetson modules, intelligent machine OEMs may now scale their product offerings with pin-compatible Jetson modules.
While containerised deployments offer flexible and seamless upgrades, pre-trained AI models from NVIDIA NGC™, coupled with the NVIDIA Transfer Learning Toolkit, enable a quicker route to inference with optimised AI networks.
NVIDIA JetPack™ SDK lets you create AI apps for Jetson TX2 NX using accelerated libraries supporting all main AI frameworks, as well as computer vision, graphics, multimedia, and more. JetPack guarantees quick time to market and lower development expenses when combined with the most recent NVIDIA tools for application creation and optimisation.
The Jetson TX2 NX is the perfect mass-market AI product platform to bring to market and keep updating over the life of a product given its ease of development and quick deployment as well as its unusual mix of form-factor, performance, and power benefit.
900-13701-0040-000
• 1792-core NVIDIA Ampere architecture GPU with 56 tensor cores
• 2x NVDLA v2.0
• 8-core Arm Cortex-A78AE v8.2 64-bit CPU 2MB L2 + 4MB L3
• 32GB 256-bit LPDDR5
• 64GB eMMC 5.1
• PVA v2.0
• Voltage input 5V, 7V-20V
• Module Power: 15W - 4
With 15W–40W power, NVIDIA Jetson AGX Orin 32GB module delivers 200 TOPS of AI performance. This provides more than 6X the performance of Jetson AGX Xavier in a compact form factor for robotics and autonomous machines.
These system-on-modules support various AI application pipelines with an NVIDIA Ampere architecture GPU, next-generation deep learning and vision accelerators, high-speed IO, and fast memory bandwidth. You can now handle natural language understanding, 3D perception, and multi-sensor fusion challenges using your largest and most complicated AI models.
Jetson is used by use case-specific application frameworks e.g., Riva for conversational AI, DeepStream for visual AI, and NVIDIA Isaac for robotics. For synthetic data generation (SDG) and fine-tuning NGC catalogue pretrained AI models, NVIDIA TAO Toolkit and NVIDIA Omniverse Replicator are time-saving tools.
NVIDIA Jetson AGX Orin Developer Kit: Getting Started
NVIDIA Jetson AGX Orin: Next-level AI Performance for Next-Gen Robotic Applications
900-13701-0050-000
• 275 TOPS
• 2048-core NVIDIA Ampere architecture
• GPU with 64 tensor cores
• 2x NVDLA v2.0
• 12-core Arm Cortex-A78AE v8.2
• 64GB 256-bit LPDDR5
• 64GB eMMC 5.1
• PVA v2.0
• Dimensions: 100mm x 87mm
• 699 pin connector
• Module Power: 15W - 60W
• Production Lifecycle: 7 years until 2031
Power configurable between 15W and 60W allows the NVIDIA Jetson AGX Orin 64GB module to provide up to 275 TOPS of AI output. In the same small form-factor for robotics and other autonomous machine use cases, this offers over 8X the performance of Jetson AGX Xavier.
With the NVIDIA Ampere GPU architecture, next-generation deep learning and vision accelerators, high-speed IO, and fast memory bandwidth, this system-on-module enables several concurrent AI application pipelines. You may now create solutions for issues like natural language comprehension, 3D perception, and multi-sensor fusion utilising your most sophisticated and largest artificial intelligence models.
Jetson maintains the NVIDIA AI software stack, besides offering user-case specific Application frameworks, including Isaac for robotics. DeepStream for visual AI, and Riva for conversational AI. The Jetson Orin provides time-saving features including NVIDIA TAO to fine-tune pretrained AI models from the NGC catalogue and NVIDIA Omniverse Replicator for synthetic data generation (SDG).
Unmatched in performance and efficiency for robots and other autonomous machines, Jetson Orin modules offer you the freedom to design the next generation of artificial intelligence solutions using the newest NVIDIA GPU technology. Your journey to market has never been quicker than with the world-standard NVIDIA AI software stack and an ecosystem of products and services.
Edge Generative AI powered by NVIDIA Jetson O
900-13701-0080-000
• 248 TOPS
• 2048-core NVIDIA Ampere architecture
• GPU with 64 Tensor Cores
• 2x NVDLA v2.0
• 12-core Arm Cortex-A78AE v8.2
• 64GB 256-bit LPDDR5 @ 3200MHz 204.8GB/s (+ECC)
• 64GB eMMC 5.1
• PVA v2.0
• Mechanical Shock: 50G Operational
• Vibration: 5G Operational
• Extended Temp. Range: -40 to +85 C
• Dimensions: 100mm x 87mm
• 699 pin connector
• Module Power: 15W – 75W
• Production Lifecycle: 10 years until 2033
Edge AI is increasingly being employed in agriculture, construction, energy, aerospace, satellites, the public sector, and other industries. You can deploy AI and computation for sensor fusion in these complicated environments with the NVIDIA Jetson Edge AI and robotics platform. The NVIDIA Jetson AGX Orin Industrial 64GB module offers AI performance up to 275 TOPS with customizable power between 15W and 60W. This provides for robotics and other autonomous machine use cases more than 6X the performance of Jetson AGX Xavier in the same small form-factor.
With an NVIDIA Ampere architecture GPU, next-generation deep learning and vision accelerators, high-speed IO, and quick memory bandwidth, these system-on-modules support many concurrent AI application pipelines. You may now create solutions utilizing your most intricate and expansive AI models to address issues like multi-sensor fusion, 3D perception, and natural language understanding.
The European Space Agency and the Barcelona Supercomputing Centre collaborated to investigate the impact of radiation on the Jetson AGX Xavier Industrial modules. More information may be found in the document Sources of Single Event Effects in the NVIDIA Xavier SoC Family Under Proton Irradiation. Their radiation statistics show that the Jetson AGX Xavier Industrial module, in conjunction with a tough casing, is an excellent choice for high-performance processing in thermally limited satellites deployed in both low-Earth and geosynchronous orbits.
Edge Generative AI powered by NVIDIA Jetson Orin
• 1024-core NVIDIA Ampere architecture
• GPU with 32 Tensor Cores
• 2x NVDLA v2.0
• 6-core Arm Cortex-A78AE v8.2 64-bit CPU 1,5MB L2 + 4MB L3
• 16GB 128-bit LPDDR5
• PVA v2.0
• Supports external NVMe
With power configurable between 10W and 40W, NVIDIA Jetson Orin 16GB series modules provide up to 157 TOPS of AI performance in the smallest Jetson form-factor. Ideal for compact formfactor, low-power items like drones and handheld gadgets, this offers 5X the performance of NVIDIA Jetson AGX Xavier and 7.5X the performance of Jetson Xavier NX.
With an NVIDIA Ampere architecture GPU, next-generation deep learning and vision accelerators, high-speed IO, and rapid memory bandwidth, these system-on-modules support several concurrent AI application pipelines. You may now create solutions for issues like natural language understanding, 3D perception, and multi-sensor fusion utilising your largest and most sophisticated artificial intelligence models.
Running the NVIDIA AI software stack, Jetson offers use-case-specific application frameworks include NVIDIA Isaac for robotics, NVIDIA Metropolis for visual AI, and NVIDIA Holoscan for sensor processing. Running the NVIDIA AI software stack, Jetson offers use-case-specific application frameworks include NVIDIA Isaac for robotics, NVIDIA Metropolis for visual AI, and NVIDIA Holoscan for sensor processing.
Easier to create, control, and deploy AI at the edge, the Jetson Xavier NX 16GB module makes use of enhanced cloud-native software features spread over the Jetson platform range.
• 1024-core NVIDIA Ampere GPU architecture with 32 tensor cores
• 1x NVDLA v2.0
• 6-core Arm® Cortex®-A78AE v8.2 64-bit CPU
• 8 GB LPDDR5
• PVA v2.0
• Supports NVMe
THE MOST SOPHISTICATED AI COMPUTER FOR SMALLER, LOWER-POWER AUTONOMOUS SYSTEMS
With power configurable between 10W and 40W, NVIDIA Jetson Orin NX 8GB series modules provide up to 117 TOPS of AI performance in the smallest Jetson form-factor. Ideal for compact form-factor, low-power products like drones and handheld gadgets, this offers 5X the performance of NVIDIA Jetson AGX Xavier and 7.5X the performance of Jetson Xavier NX.
Thanks to the NVIDIA Ampere architecture GPU, next-generation deep learning and vision accelerators, high-speed I/O, and fast memory bandwidth, these system-on-modules can run multiple AI application processes at simultaneously. thanks to an NVIDIA Ampere architecture GPU, next-generation deep learning and vision accelerators, high-speed I/O, and fast memory bandwidth, these system-on-modules can run multiple AI application processes at simultaneously. time. Now you can use your most sophisticated and most complex AI models to assist with answers to problems like understanding natural language, observing things in 3D, and combining data from multiple sensors.
Running the NVIDIA AI software stack, Jetson offers use-case-specific application frameworks include NVIDIA Isaac for robotics, NVIDIA Metropolis for visual AI, and NVIDIA Holoscan for sensor processing. Running the NVIDIA AI software stack, Jetson offers use-case-specific application frameworks include NVIDIA Isaac for robotics, NVIDIA Metropolis for visual AI, and NVIDIA Holoscan for sensor processing.
When it comes to robotics and other autonomous machines, Jetson Orin modules offer unrivalled performance and efficiency, and they also give you the freedom to develop the next generation of AI solutions using cutting-edge NVIDIA GPU technology.
Edge Generative AI powered by NVIDIA Jetson Orin
900-13767-0030-000
• 1024-core NVIDIA Ampere
• architecture GPU with 32 tensor cores
• 6-core Arm Cortex-A78AE v8.2
• 8GB 128-bit LPDDR5
• Support for external NVMe
With power settings that can range from 7W to 25W, the NVIDIA Jetson Orin Nano 8GB module gives you up to 67 TOPS of AI speed in the smallest Jetson form factor. This makes the NVIDIA Jetson TX2 NX up to 50 times faster and extends the range of entry-level Edge AI.
In addition to being pin-compatible with Jetson Orin NX, this system-on-module supports multiple AI application streams at the same time, has fast I/O, and houses an NVIDIA Ampere architecture GPU. Therefore, cheaper consumer, retail analytics, and industrial quality assurance (QA) devices can use more complicated AI models.
The NVIDIA AI software stack is run by Jetson. It has application libraries for different use cases, such as NVIDIA Isaac for robotics, NVIDIA Metropolis for vision AI, and NVIDIA Holoscan for sensor processing. When you use NVIDIA Omniverse Replicator to make synthetic data (SDG) and NVIDIA TAO Toolkit to fine-tune AI models that have already been learnt from the NGC™ catalogue, you can save a lot of time.
When it comes to speed and efficiency, Jetson Orin modules are among the finest for robots and other self-driving machines. With the latest NVIDIA GPU technology, they also give you the freedom to make the next generation of AI solutions. When you use the world-renowned NVIDIA AI software stack and a network of services and goods, you can get to market faster than ever.
Visit https://developer.nvidia.com/jetson to find out more.
Edge Generative AI powered by NVIDIA Jetson Orin
With power adjustable from 7W to 25W, the NVIDIA Jetson Orin Nano 4GB module offers up to 34 TOPS of AI performance in the smallest Jetson form-factor. This establishes the new baseline for entry-level Edge AI and gives you up to 68X the performance of NVIDIA Jetson Nano.
This system-on-module is pin-compatible with Jetson Orin NX, and supports multiple concurrent AI application pipelines with high-speed IO and an NVIDIA Ampere architecture GPU. This enables lower-priced consumer, retail analytics, and industrial QA devices to benefit from more complex AI models.
Running the NVIDIA AI software stack, Jetson offers use-case-specific application frameworks include NVIDIA Isaac for robotics, NVIDIA Metropolis for visual AI, and NVIDIA Holoscan for sensor processing. NVIDIA TAO Toolkit for fine-tuning pretrained AI models from the NGC catalogue and NVIDIA Omniverse Replicator for synthetic data generation (SDG) help you to save great time.
Unmatched in performance and efficiency for robots and other autonomous machines, Jetson Orin modules offer you the freedom to design the next generation of AI solutions using the newest NVIDIA GPU technology. Your journey to market has never been quicker than with the world-standard NVIDIA AI software stack and an ecosystem of products and services.
Visit https://developer.nvidia.com/jetson to find out more.
Edge Generative AI powered by NVIDIA Jetson Orin
900-82888-0040-000
• 512-Core NVIDIA Volta GPU with Tensor Cores
• (2x) NVDLA Engines
• 8-Core ARM® v8.2 64-Bit Carmel CPU
• 32GB 256-Bit LPDDR4x|136.5GB/s
• 32 GB eMMC 5.1 Flash Storage
• 7-Way VLIW Vision Accelerator Processor
In a small package using under 30 watts, the NVIDIA Jetson AGX Xavier provides up to 32TOPS of accelerated processing power.
Designed for performance-driven and cost-effective autonomous machine applications, this sophisticated system-on-module runs on the NVIDIA Xavier SoC. Heterogeneous accelerated computing architecture enables artificial intelligence at the edge and offers improved compute performance. Its creative thermal design, integrated memory, storage, power management, and other features help to accelerate time to market. Visual inspection, production, logistics, smart cities, healthcare, retail, and services are just a few of the areas where Jetson AGX Xavier can handle current AI workloads.
Software libraries for deep learning, computer vision, GPU computing, multimedia processing, and much more are included in the NVIDIA JetPack SDK, which supports Jetson AGX Xavier. The SDK also includes Linux OS, NVIDIA CUDA, cuDNN, and TensorRT. Both the NVIDIA Isaac SDK, which provides software tools for developing robots, and the NVIDIA DeepStream SDK, which provides realtime situational awareness through intelligent video analytics (IVA), are compatible with it. In addition to lowering development costs and effort, these aid in speeding up software development and improving performance.
Download the SDK from the NVIDIA Developer Site and locate all you require to begin. Running autonomous machines software, faster and with less power, NVIDIA Jetson systems offer the performance and power economy. Each is a full System-on-Module (SOM) with flash storage, PMIC, DRAM, CPU, and GPU, thereby saving development time and cost. Jetson is extendable as well. Simply choose the SOM appropriate for the application and construct the bespoke system around it to satisfy its particular requirements.
900-82888-0050-000
• 512-Core NVIDIA Volta GPU with 64 Tensor Cores 22 TOPS (INT8)
• (2x) NVDLA Engines 10 TOPS [INT8]
• 8-Core ARM® v8.2 64-Bit Carmel CPU 8MB L2 + 4MB L3
• 64GB 256-Bit LPDDR4x|136.5GB/s
• 32 GB eMMC 5.1 Flash Storage
• 7-Way VLIW Vision Accelerator Processor
Running in as little as 10W and producing up to 32 TOPs, the Jetson AGX Xavier 64GB module enables AI-powered autonomous devices. The 64GB RAM lets users save several AI models, execute sophisticated apps, and improve their real-time pipelines. Being part of the top AI computing platform in the world, it gains from the extensive collection of NVIDIA AI tools and processes, which helps developers to rapidly train and deploy neural networks. NVIDIA JetPack SDK supports Jetson AGX Xavier 64GB, which can help you reduce development work and cost.
Modern AI workloads can be executed on the Jetson AGX Xavier, which can also address issues in optical inspection, manufacturing, robotics, logistics, retail, service, agriculture, smart cities, and healthcare. The NVIDIA JetPack SDK, which comprises board support package (BSP), Linux OS, NVIDIA CUDA®, cuDNN, and TensorRT software libraries for deep learning, computer vision, GPU computing, multimedia processing, and much more, supports Jetson AGX Xavier. The NVIDIA DeepStream SDK additionally supports it by means of intelligent video analytics (IVA) and NVIDIA Isaac SDK, which provides a software framework for robot development, hence providing a whole toolset for real-time situational awareness. While lowering development cost and effort, these help to improve performance and speed up software creation.
Visit https://developer.nvidia.com/jetson to find out more.
Download the SDK from the NVIDIA Developer Site and locate all you require to begin. Running autonomous machines software, faster and with less power, NVIDIA Jetson systems offer the performance and power economy. Each is a full System-on-Module (SOM) with flash storage, PMIC, DRAM, CPU, and GPU, thereby saving development time and cost. Jetson is extendable as well. Simply choose the SOM appropriate for the application and construct the bespoke system around it to satisfy its specific requirements.
900-82888-0080-000
• 512-Core NVIDIA Volta GPU with Tensor Cores
• (2x) NVDLA Engines
• 8-Core ARM® v8.2 64-Bit Carmel CPU
• 32GB 256-Bit LPDDR4x|136.5GB/s
• 32 GB eMMC 5.1 Flash Storage
• 7-Way VLIW Vision Accelerator Processor
In a power-efficient, tough form-factor, NVIDIA Jetson AGX Xavier Industrial offers the maximum performance for AI integrated industrial and functional safety applications. This system-on-module enables consumers bring cutting-edge AI models to their most demanding use cases with up to 20X the performance and 4X the memory of Jetson TX2i. Perfect for industrial grade and/or safety certified AI products, the Jetson AGX Xavier Industrial module offers enhanced temperature range, shock, and vibration specifications as well as new functional safety features supervised by the Safety Cluster Engine. Other security measures to safeguard customer software included in the module are hardware validated secure boot, hardware accelerated cryptography, support for encrypted storage and memory.
You can reduce development time and costs for AI software for production with the help of the NVIDIA JetPack SDK and pretrained models from the NGC catalogue. Included in JetPack are the most recent software development tools from NVIDIA, the Jetson Linux driver package (L4T), and libraries that accelerate all major AI frameworks, graphics, multimedia, computer vision, and more. Over the course of the industrial module's 10-year lifespan, cloud-native technologies provide smooth model and software upgrades, and NGC's pre-trained models are performance optimised and prepared to be fine-tuned with customer datasets.
Customers interested in developing products that comply with IEC 61508 standards and using Jetson AGX Xavier Industrial for both safety-related and non-safety applications can purchase the Jetson Safety Extension Package, which includes the necessary software and documentation. For customers looking for examples of ISO13849 Cat. 2/3 PLd architecture as well as recommendations, guidelines, and other useful information, we have a safety design guide accessible. With its form-factor, performance, and power advantages, as well as its ease of development and rapid deployment, Jetson AGX Xavier Industrial is the perfect platform for bringing robust and safety-ready AI products to market and for continuous updates throughout their lifetime.
900-83310-0001-000
The most innovative technology for AI computing and visual computing comes in a supercomputer the size of a credit card. Its small form factor and power envelope make the Jetson TX2 module ideal for intelligent edge devices like robots, drones, smart cameras, and portable medical devices.
• NVIDIA Pascal architecture GPU
• Dual-core Denver 2 64-bit CPU + quad-core
• ARM A57 Complex
• 8 GB L128 bit LPDDR4
• 32 GB eMMC 5.1
Supercomputers the size of credit cards have the most cutting-edge technology for artificial intelligence and visual computing. The Jetson TX2 module is perfect for smart cameras, drones, robotics, and portable medical equipment because to its small size and power envelope.
The Jetson TX2 is easily integrated into a wide range of products and form factors thanks to its choice of common hardware connectors. It also contains the full Jetpack software development kit (SDK), which can speed up your program development with tools like the BSP and libraries for deep learning, computer vision, graphics processing units (GPUs), multimedia processing, and more.
Technical note: When used with any of the Connect Tech carrier boards: CTI L4T needs to be loaded for full functionality: (Further info)
900-83489-0000-000
• NVIDIA Pascal architecture GPU
• Dual-core Denver 2 64-bit CPU and quad-core
• Arm A57 Complex
• 8 GB 128 bit LPDDr4 (ECC support)
• 32 GB eMMC 5.1
Jetson TX2i has several common hardware interfaces that facilitate its integration into a broad spectrum of products and form factors. In addition, it includes the whole NVIDIA Jetpack SDK, which has the BSP, deep learning libraries, computer vision, GPU computing, multimedia processing, and much more to speed your software creation.
EXTEND AI PERFORMANCE TO INDUSTRIAL ENVIRONMENTS WITH THE HIGH-PERFORMANCE, LOW-POWER JETSON TX2I..
Jetson TX2i features a variety of standard hardware interfaces that make it easy to integrate it into a wide range of products and form factors. Plus, it comes with the complete NVIDIA Jetpack SDK, which includes the BSP, libraries for deep learning, computer vision, GPU computing, multimedia processing, and much more to accelerate your software development.
A supercomputer the size of a credit card holds the most creative technologies for GPU-enabled AI processing. High-performance edge computing devices including industrial robots, machine vision cameras, and portable medical equipment are best suited for the NVIDIA Jetson TX2i module because of its sturdy build, compact size, and lower power envelope.
Built for consistent performance in demanding industrial settings, Jetson TX2i has long operating life (MTBF) and comes with an extended warranty and sales lifecycle.
Technical note: When used with any of the Connect Tech carrier boards: CTI L4T needs to be loaded for full functionality: (Further info)
• 256-core Pascal™ GPU
• Dual-core NVIDIA Denver 2 Quad-core ARM A57 complex
• ARM A57 Complex
• 4 GB 128 Bit LPDDR4 | 25.6 GB/s
• 16 GB eMMC 5.1
Running contemporary AI applications, the NVIDIA Jetson TX2 4GB offers strong computational performance in a tiny credit card shape. Configurable between 7.5W and 15W, the module is perfect for a variety of intelligent edge devices like robots, drones, smart cameras, and portable medical devices.
At the same price, Jetson TX2 4GB provides double the performance per watt of Jetson TX1, hence allowing the creation of inexpensive, energy efficient, high-performance products. From low speed GPIOs to high-speed PCIe and USBs, Jetson TX2 4GB boasts several conventional hardware connections. This facilitates integration into a broad spectrum of form factors and products. NVIDIA JetPack also supports the module; it comprises a board support package (BSP), Linux OS, NVIDIA CUDA, cuDNN, and TensorRT™ software libraries for deep learning, computer vision, GPU computing, multimedia processing, and much more.
The software is simple to use and allows you to get started quickly. The same Jetpack SDK is used throughout the NVIDIA Jetson family of products and is completely interoperable with NVIDIA's world-class AI platform for training and deploying AI software. This tried-and-true software stack simplifies development and saves time and effort.
Technical note: When used with any of the Connect Tech carrier boards: CTI L4T needs to be loaded for full functionality: (Further info)
900-83668-0000-000
Jetson Xavier NX System-on-Module (SOM):
• 384-Core Volta GPU
• 6-Core ARM 64 Bit CPU
• 8 GB LPDDR4
• 16 GB eMMC
The NVIDIA Jetson Xavier NX is a compact system-on-module (SOM) that is smaller than a credit card that delivers supercomputer capability. It has new cloud-native capability and outperforms its widely accepted predecessor, Jetson TX2, by more than 10 times. This power efficiency enables precise multi-modal AI inference in a tiny form factor, paving the way for revolutionary edge devices in manufacturing, logistics, retail, service, agriculture, smart cities, healthcare, and life sciences, and more.
The Jetson Xavier NX module now has cloud-native support throughout the Jetson platform line - up, making it easier to build, deploy, and manage AI at the edge. Pre-trained AI models from NVIDIA NGC, in conjunction with the NVIDIA Transfer Learning Toolkit, enable a faster path to inference with optimized AI networks, while containerized deployment to Jetson devices enables flexible and seamless updates.
The NVIDIA JetPack SDK enables the creation of AI applications for the Jetson Xavier NX by providing accelerated libraries that support all major AI frameworks, as well as computer vision, graphics, multimedia, and other technologies. JetPack, in conjunction with the most recent NVIDIA tools for application development and optimization, ensures a quick time to market and lower development costs. Jetson Xavier NX is the most flexible and scalable platform to get to market and continuously update over the lifetime of a product due to its ease of development and speed of deployment, as well as a unique combination of form-factor, performance, and power advantage.
Announcing Cloud-Native Support for Jetson Platform
900-83668-0030-000
Jetson Xavier NX System-on-Module (SOM):
• 384-Core Volta GPU
• 6-Core ARM 64 Bit CPU
• 16 GB 128-bit LPDDR4
• 16 GB eMMC4
The NVIDIA® Jetson Xavier™ NX 16GB module benefits from new cloud-native software capabilities across the entire Jetson platform line up, making it easier to build, manage, and deploy AI at the edge. NVIDIA® Jetson Xavier™ NX 16GB brings supercomputer performance to the edge in a compact system-on-module (SOM) that is smaller than a credit card. The energy-efficient Jetson Xavier NX module delivers server-class performance—up to 14 TOPS at 10W or 21 TOPS at 15W or 20W. This unique combination of form-factor, performance, and power advantage opens the door for innovative edge devices in the fields of manufacturing, logistics, retail, service, agriculture, smart city and medical instruments. The Jetson Xavier NX 16GB module benefits from new cloud-native software capabilities across the entire Jetson platform line up, making it easier to build, manage, and deploy AI at the edge.
Pre-trained AI models from NVIDIA NGC together with the NVIDIA TAO Toolkit give a faster path to trained and optimized AI networks. Containerized deployment to Jetson devices allows flexible and seamless updates.
NVIDIA JetPack™ SDK enables multi-modal AI application development for Jetson Xavier NX 16GB with accelerated libraries supporting all major AI frameworks, as well as computer vision, computer graphics, multimedia, and more. Together with the latest NVIDIA tools for application development and optimization, JetPack™ ensures fast time to market and reduced development costs.
Designed for ease of development and speed of deployment, Jetson is the most flexible and scalable platform to get to market and continuously update AI software over the lifetime of a product.
945-13730-0055-000
• 2048-core NVIDIA Ampere architecture GPU with 64 Tensor cores
• 2x NVDLA v2.0
• 12-core Arm Cortex-A78AE v8.2 64-bit CPU
• 64GB 256-bit LPDDR5
• 64GB eMMC 5.1
• PVA v2.0
Getting started with Jetson Orin is made simple thanks to the NVIDIA Jetson AGX Orin Developer Kit. This development kit is great for creating sophisticated AI-powered robots and other autonomous machines given its small size, many ports, and up to 275 TOPS of AI performance. The kit includes a Jetson AGX Orin module that can emulate all Jetson Orin modules. The NVIDIA Ampere GPU architecture, next-generation deep learning and vision accelerators, high-speed IO, and rapid memory bandwidth enable it to serve several AI application pipelines at the same time. You can now use your most powerful and complicated AI models to solve challenges like natural language understanding, 3D perception, and multi-sensor fusion.
Jetson manages the NVIDIA AI software stack, and use-case-specific application frameworks are available, such as Isaac for robotics, DeepStream for vision AI, and Riva for conversational AI. You can save time by using NVIDIA Omniverse Replicator for synthetic data generation (SDG) and NVIDIA TAO toolbox to fine-tune pretrained AI models from the NGC™ catalogue.
With the computing power of more than 8 Jetson AGX Xavier computers in a development kit that combines the latest NVIDIA GPU technology with the world's most comprehensive deep learning software stack, you'll be able to create both tomorrow's and today's AI solutions.
945-13766-0005-000
• NVIDIA Ampere architecture with 1024 NVIDIA CUDA cores with 32 tensor cores
• 6-core Arm Cortex-A78AE v8.2 64-bit CPU
• 8GB 128-bit LPDDR5 102 GB/s
• Supports for external via microSD slot
• external NVMe via M.2 KeyM
The NVIDIA Jetson Orin Nano Super Developer Kit is a small yet powerful computer that reimagines generative AI for small edge devices. It provides up to 67 TOPS of AI performance a 1.7X increase over its predecessor and can run a wide range of generative AI models, including vision transformers, big language models, vision-language models, and more. It offers developers, students, and makers the most affordable and accessible platform, backed by NVIDIA AI software and a vast AI software ecosystem, to democratise generative AI at the edge. Existing Jetson Orin Nano Developer Kit customers can benefit from this performance gain by simply upgrading their software, allowing everyone to explore new possibilities with generative AI.
The developer kit includes a Jetson Orin Nano 8GB module and a reference carrier board that can hold all Orin Nano and Orin NX modules, making it an ideal foundation for prototyping your next-generation edge AI product. The Jetson Orin Nano 8GB module includes an Ampere GPU and a 6-core ARM CPU, allowing for several concurrent AI application pipelines and high-performance inference. The carrier board has a large number of ports, including two MIPI CSI connectors that accommodate camera modules with up to four lanes, allowing for greater resolution and frame rate than before.
The world's most affordable generative AI computer
Generative AI Models at the Edge Powered by NVIDIA Jetson Orin
945-14070-0085-000
• Includes a Quick Start Guide
• NVIDIA Jetson T5000 module with reference carrier board and heat sink, DC power 140W,
• 802.11ax wireless network interface controller
• 1 TB NVMe populated in M.2 Key-M slot
The Most Innovative Humanoid Robotics Platform
The best platform for humanoid robotics is the NVIDIA Jetson AGX Thor Developer Kit, which offers unparalleled scalability and performance. With 128 GB of RAM and an NVIDIA Blackwell GPU, it can easily execute the newest generative AI models with up to 2070 FP4 TFLOPS of AI compute all within a 130W power envelope. It offers up to 7.5x more AI computation and 3.5x more energy efficiency than the NVIDIA Jetson AGX Orin.
Jetson Thor's powerful 14-core Arm Neoverse-V3AE CPU and innovative Blackwell Multi-Instance GPU (MIG) technology enable you speed up low-latency, real-time applications. A third-generation Programmable Vision Accelerator (PVA), two encoders and decoders, an optical flow accelerator, and other accelerators are also included. The developer kit provides a wide range of I/O options for high-speed sensor fusion, such as a wired Multi-GbE RJ45 connector, several USB ports, a QSFP slot with 4x25GbE, and various connectivity interfaces. Additionally, it is built to integrate seamlessly with current humanoid robot platforms, making tethering simple and enabling rapid prototyping.
Designed from the ground up to power the next generation of humanoid robots, Jetson Thor is a member of a new class of robotic computers. From all of the well-known LLMs and VLMs to Vision Language Action (VLA) models like the NVIDIA Isaac GR00T N, it supports a broad variety of generative AI models. Jetson Thor uses the NVIDIA AI software stack for physical AI applications, such as NVIDIA Holoscan for sensor processing, NVIDIA Metropolis for visual agentic AI, and NVIDIA Isaac for robotics, to provide a smooth cloud-to-edge experience. NVIDIA agentic AI procedures, such as Video Search and Summarisation (VSS), can also be used to create AI agents at the edge.
To speed up solution creation in sectors including robotics, smart spaces, retail, industrial, medical, and more, our network of partners provides all the carrier boards, design services, cameras, and other sensors you require in addition to extra AI and system software.
The Jetson AGX Thor Developer Kit is the perfect platform for developers wishing to explore new possibilities for physical AI applications, such as humanoid robotics, because of its massive AI performance and sensor capabilities in a small form factor.
Authorized Distributor / Reseller
The Junction AS
+47 40 40 34 02
Sales@TheJunction.eu
https://www.TheJunction.eu
DISCLAIMER
The content of this interactive catalogue is protected by copyright. Unauthorized use may be subject to criminal prosecution. The Junction accepts no liability for any of the publicly available information contained within this publication.
© 2015-2025 The Junction.eu - Drammen, Norway | Århus, Denmark