NVIDIA Jetson advanced AI system: build autonomous machines for manufacturing, retail, robotics, smart farming, and healthcare 

Buy the latest NVIDIA Jetson Production module powered edge vices, carrier boards, and add-ons at Seeed

NVIDIA Jetson compatible Edge AI platform at Seeed
Carrier board and add-ons for NVIDIA Jetson platform

About NVIDIA Jetson Modules 

NVIDIA® Jetson™ brings accelerated AI performance to the Edge in a power-efficient and compact form factor. The Jetson family of modules all use the same NVIDIA CUDA-X™ software, and support cloud-native technologies like containerization and orchestration to build, deploy, and manage AI at the edge. Together with NVIDIA JetPack™ SDK, Jetson modules will help you develop and deploy innovative products across all industries.

Jetson Module and devices Lineup

Jetson AGX Orin Series

📷 NVIDIA: Jetson AGX Orin Series

AI Performance200 TOPS (INT8)275 TOPS (INT8)
GPUNVIDIA Ampere architecture with 1792 NVIDIA® CUDA® cores and 56 Tensor CoresNVIDIA Ampere architecture with 2048 NVIDIA® CUDA® cores and 64 Tensor Cores
Max GPU Freq939 MHz1.3 GHz
CPU8-core Arm® Cortex®-A78AEv8.2 64-bit CPU2MB L2 + 4MB L312-core Arm® Cortex®-A78AEv8.2 64-bit CPU3MB L2 + 6MB L3
CPU Max Freq2.2 GHz
DL Accelerator2x NVDLA v2.0
DLA Max Freq1.4 GHz1.6 GHz
Vision AcceleratorPVA v2.0
Memory32GB 256-bit LPDDR5204.8 GB/s64GB 256-bit LPDDR5204.8 GB/s
Storage64GB eMMC 5.1
CSI CameraUp to 6 cameras (16 via virtual channels*)16 lanes MIPI CSI-2D-PHY 2.1 (up to 40Gbps) | C-PHY 2.0 (up to 164Gbps)
Video Encode1x 4K60 | 3x 4K30 | 6x 1080p60 |12x 1080p30 (H.265)H.264, AV12x 4K60 | 4x 4K30 | 8x 1080p60 |16x 1080p30 (H.265)H.264, AV1
Video Decode1x 8K30 | 2x 4K60 | 4x 4K30 | 9x 1080p60|18x 1080p30 (H.265)H.264, VP9, AV11x 8K30 | 3x 4K60 | 7x 4K30 | 11x 1080p60|22x 1080p30 (H.265)H.264, VP9, AV1
UPHYUp to 2 x8, 1 x4, 2 x1 (PCIe Gen4, Root Port & Endpoint)3x USB 3.2Single lane UFS
Networking1x GbE4x 10GbE
Display1x 8K60 multi-mode DP 1.4a (+MST)/eDP 1.4a/HDMI 2.1
Other I/O4x USB 2.04x UART, 3x SPI, 4x I2S, 8x I2C, 2x CAN, DMIC & DSPK, GPIOs
Power15W – 40W15W – 60W
Mechanical100mm x 87mm699-pin Molex Mirror Mezz ConnectorIntegrated Thermal Transfer Plate

Jetson AGX Orin Dev Kit 

Bring your next-gen products to life with the world’s most powerful AI computer for energy-efficient autonomous machines. Up to 275 TOPS and 8X the performance of the last generation for multiple concurrent AI inference pipelines and high-speed interface support for multiple sensors make this the ideal solution for applications from manufacturing and logistics to retail and healthcare.

reServer with Jetson AGX Orin 64GB

reServer Jetson is a powerful inference server on the edge powered by Jetson AGX ORIN 64GB, which delivers up to 275 TOPs AI performance. It provides a high-speed 10-Gigabit Ethernet port and supports hybrid connectivity including 5G, LoRa, BLE and WiFi. reServer Jetson with pre-installed Triton Inference server is ready for fast and scalable AI in production and concurrent model execution. 

Also check out our blog Compare NVIDIA Jetson AGX Orin with AGX Xavier: 8x AI performance, in-advance Ampere GPU, CPU, Memory & Storage

Jetson Orin NX Series

📷 NVIDIA: Jetson Orin NX module

Jetson Orin NX modules deliver up to 100 TOPS of AI performance in the smallest Jetson form factor, with power configurable between 10W and 25W. The Orin NX Series is form factor compatible with the Jetson Xavier NX series and delivers up to 5x the performance, or up to 3X the performance at the same price. 

Jetson Xavier NX SeriesJetson Orin NX Series
Jetson Xavier NX 16GBJetson Xavier NXJetson Orin NX 8GBJetson Orin NX 16GB
AI Performance21 TOPs21 TOPs70 TOPS100 TOPS
GPU384-core NVIDIA Volta™ GPU with 48 Tensor Cores384-core NVIDIA Volta™ GPU with 48 Tensor Cores1024-core NVIDIA Ampere GPU with
32 Tensor Cores
1024-core NVIDIA Ampere GPU with
32 Tensor Cores
CPU6-core NVIDIA Carmel Arm®v8.2 64-bit CPU
6MB L2 + 4MB L3
6-core NVIDIA Carmel Arm®v8.2 64-bit CPU
6MB L2 + 4MB L3
6-core NVIDIA Arm® Cortex A78AE v8.2 64-bit CPU
1.5MB L2 + 4MB L3
8-core NVIDIA Arm® Cortex A78AE v8.2 64-bit CPU
2MB L2 + 4MB L3
DL Accelerator2x NVDLA v12x NVDLA v11x NVDLA v2.02x NVDLA v2.0
Memory16 GB8 GB8 GB16 GB
Storage16 GB eMMC 5.1 *16 GB eMMC 5.1 *
(Supports external NVMe)

(Supports external NVMe)
Power10W | 15W | 20W10W | 15W | 20W10W – 20W10W – 25W
Mechanical69.6 mm x 45 mm260-pin SO-DIMM connector69.6 mm x 45 mm 260-pin SO-DIMM connector69.6mm x 45mm 260-pin SO-DIMM connector69.6mm x 45mm 260-pin SO-DIMM connector

Jetson AGX Xavier

📷 NVIDIA: Jetson AGX Xavier Module
AI Performance32 TOPs
GPU512-core Volta GPU with 64 Tensor Cores11 TFLOPS (FP16)22 TOPS (INT8)
CPU8-core Carmel ARM v8.2 64-bit CPU, 8MB L2 + 4MB L3
Memory32GB 256-Bit LPDDR4x | 136.5GB/s
Storage32GB eMMC 5.1
DL Accelerator(2x) NVDLA Engines* | 5 TFLOPS (FP16), 10 TOPS (INT8)
Vision Accelerator7-way VLIW Vision Processor*
Video Encode2x1000MP/sec4x 4K @ 60 (HEVC)8x 4K @ 30 (HEVC)16x 1080p @ 60 (HEVC)32x 1080p @ 30 (HEVC)
Video Decode2x1500MP/sec2x 8K @ 30 (HEVC)6x 4K @ 60 (HEVC)12x 4K @ 30 (HEVC)26x 1080p @ 60 (HEVC)52x 1080p @ 30 (HEVC)30x 1080p @ 30 (H.264)
Size105 mm x 105 mm

Jetson Xavier AGX H01 Kit 

The Jetson Xavier AGX H01 Development Kit is powered by NVIDIA Jetson AGX Xavier processor which applies AI performance and delivers up to 32 Tera Operations Per Second(TOPs) yet costs less than 30W. It is an ideal solution to agriculture, optical inspection, manufacturing, robotics, logistics, retail, service, smart cities, and healthcare.

📷 Seeed: Jetson Xavier AGX H01 Kit

Jetson Xavier NX 

Xavier NX can achieve 172 FPS for PeopleNet- ResNet34  of People Detection, 274 FPS for DashCamNet-ResNet18 of Vehicle Detection, and 1126 FPS for FaceDetect-IR-ResNet18 of Face Detection. Benchmark details can be found on NVIDIA®’s DeepStream SDK website.

📷 NVIDIA: Jetson Xavier NX Module
AI Performance21 TOPS (INT8)
GPU384-core NVIDIA Volta™ GPU with 48 Tensor Cores
GPU Max Freq1100 MHz
CPU6-core NVIDIA Carmel ARM®v8.2 64-bit CPU6MB L2 + 4MB L3
CPU Max Freq2-core @ 1900MHz4/6-core @ 1400Mhz
Memory8 GB 128-bit LPDDR4x @ 1866MHz, 59.7GB/s
16 GB 128-bit LPDDR4x @ 1866 MHz, 59.7GB/s
Storage16 GB eMMC 5.1
PCIe1 x1 + 1×4(PCIe Gen3, Root Port & Endpoint)
CSI CameraUp to 6 cameras (36 via virtual channels)12 lanes MIPI CSI-2D-PHY 1.2 (up to 30 Gbps)
Video Encode2x 4K60 | 4x 4K30 | 10x 1080p60 | 22x 1080p30 (H.265)2x 4K60 | 4x 4K30 | 10x 1080p60 | 20x 108p30 (H.264)
Video Decode2x 8K30 | 6x 4K60 | 12x 4K30 | 22x 1080p60 | 44x 1080p30 (H.265)2x 4K60 | 6x 4K30 | 10x 1080p60 | 22x 1080p30 (H.264)
Display2 multi-mode DP 1.4/eDP 1.4/HDMI 2.0
DL Accelerator2x NVDLA Engines
Vision Accelerator7-Way VLIW Vision Processor
Networking10/100/1000 BASE-T Ethernet
Mechanical45 mm x 69.6 mm260-pin SO-DIMM connector

Jetson Nano 

📷 NVIDIA: Jetson Nano Module
GPUNVIDIA Maxwell architecture with 128 NVIDIA CUDA® cores
CPUQuad-core ARM Cortex-A57 MPCore processor
Memory4 GB 64-bit LPDDR4, 1600MHz 25.6 GB/s
Storage16 GB eMMC 5.1
Video Encode250MP/sec1x 4K @ 30 (HEVC)2x 1080p @ 60 (HEVC)4x 1080p @ 30 (HEVC)4x 720p @ 60 (HEVC)9x 720p @ 30 (HEVC)
Video Decode500MP/sec1x 4K @ 60 (HEVC)2x 4K @ 30 (HEVC)4x 1080p @ 60 (HEVC)8x 1080p @ 30 (HEVC)9x 720p @ 60 (HEVC)
Camera12 lanes (3×4 or 4×2) MIPI CSI-2 D-PHY 1.1 (1.5 Gb/s per pair)
ConnectivityGigabit Ethernet, M.2 Key E
DisplayHDMI 2.0 and eDP 1.4
USB4x USB 3.0, USB 2.0 Micro-B
Mechanical69.6 mm x 45 mm260-pin edge connector

reComputer Nano/NX: real world AI at the Edge, starts from $199

Built with Jetson Nano 4GB/ Xavier NX 8GB/16GB

  • Edge AI box fit into anywhere
  • Embedded Jetson Nano/NX Module
  • Pre-installed Jetpack for easy deployment
  • Nearly same form factor with Jetson Developer Kits, with rich set of I/Os
  • Stackable and expandable

Jetson Benchmark: Jetson Xavier NX and Jetson AGX Orin MLPerf v2.0 Results

ModelJetson Xavier NXJetson AGX XavierJetson AGX Orin
Action Recognition 2D2454711577
Action Recognition 3D2132105
LPR Net70611904118
Dashcam Net4256711908
Bodypose Net105172559
ASR: Citrinet 10242734113
NLP: BERT-base5894287
TTS: Fastpitch-HifiGAN7942

Get started with Jetson Software with necessary ML tools

Jetson Platform includes all the necessary ML tools to accelerate your next AI product quickly to market. At the base of the software stack, JetPack SDK includes the Board Support Package (BSP), with boot loader, Linux kernel, drivers, toolchains, and a reference file system based on Ubuntu.

Jetson Software comprises JetPack SDK, TAO & Pretrained AI Models from the NVIDIA NGC™ catalog, Triton Inference Server, NVIDIA Riva, DeepStream SDK, and NVIDIA Isaac.

Jetson Software Ecosystem 📷: NVIDIA

Develop stream analytics with the NVIDIA DeepStream. 

NVIDIA’s DeepStream SDK delivers a complete streaming analytics toolkit for AI-based multi-sensor processing, video, audio and image understanding. 

DeepStream offers exceptional throughput for a wide variety of object detection, image classification and instance segmentation based AI models. To reduce development efforts and increase throughput, developers can use highly accurate pre-trained models from TAO Toolkit and deploy with DeepStream. The following table shows the end-to-end application performance from data ingestion, decoding, image processing to inference. It takes multiple 1080p/30fps streams as input. Note that running on the DLAs for Jetson Xavier NX and Jetson AGX Xavier frees up GPU for other tasks.

Jetson Nano*Jetson TX2*Jetson Xavier NXJetson AGX Xavier
ApplicationModelsInference ResolutionPrecisionModel AccuracyGPU (FPS*)GPU (FPS)GPU (FPS)DLA1 (FPS)DLA2 (FPS)GPU (FPS)DLA1 (FPS)DLA2 (FPS)
People DetectPeopleNet-ResNet34960×544INT884%123117248483055353
Vehicle DetectTrafficCamNet-ResNet18960×544INT884%19512748989486111111
Vehicle DetectDashCamNet-ResNet18960×544INT880%18462619191460116116
Face DetectFaceDetect-IR-ResNet18384×240INT896%10127611264554552007624624

Tabulated data is in FPS with 1080p/30fps input. Batch size is equivalent to the number of input streams.which is total FPS from the table divide by 30.

* FP16 inference on Jetson Nano and Tx2

Jetson AI courses and certifications 

NVIDIA’s Deep Learning Institute (DLI) delivers practical, hands-on training and certification in AI at the edge for developers, educators, students, and lifelong learners. This is a great way to get the critical AI skills you need to thrive and advance in your career. You can even earn certificates to demonstrate your understanding of Jetson and AI when you complete these free, open-source courses.

Seeed get started guide for Jetson platform 

SSD boot instruction through SDK manager or command lines

Memory Expansion for reComputer Nano

The reComputer Jetson is sold with 16 GB of eMMC and has ubuntu 18.04 LTS and NVIDIA JetPack 4.6 installed, so the remaining user space available is about 2 GB, which is a significant obstacle to using the reComputer Jetson for training and deployment in some projects. This tutorial will introduce the expansion process for different models of reComputer Jetson based on this situation and help developers to expand their systems by transferring them to external storage devices.

Leave a Reply

Your email address will not be published.


May 2022