Learn, Implement and Share

Introduction

NVIDIA Jetson AGX Xavier is an embedded system-on-module (SoM) from the NVIDIA AGX Systems family. The Jetson AGX Xavier module makes AI-powered autonomous machines possible, running in as little as 10W and delivering up to 32 TOPs. It is a computer available now for embedded designers, researchers, and DIY makers, delivering the power of modern AI in a compact, easy-to-use platform with full software programmability. It is a small, powerful computer that lets you run multiple neural networks in parallel for applications like image classification, object detection, segmentation, and speech processing. 

Jetson AGX Xavier Developer Kit
Jetson AGX Xavier Developer Kit
PlatformCPUGPUMemoryStorageMRP
Jetson AGX Xavier 8x ARM v8.2 @ 2.26GHz512x Volta GPU @ 64 Tensor Cores32GB LPDDR4 (137GB/s)32GB eMMC$699
Jetson TX14x ARM Cortex A57 @ 1.73 GHz256x Maxwell @ 998 MHz (1 TFLOP)4GB LPDDR4 (25.6 GB/s)16 GB eMMC$499
Jetson Nano4x ARM Cortex A57 @ 1.43 GHz128x Maxwell @ 921 MHz (472 GFLOPS)4GB LPDDR4 (25.6 GB/s)Micro SD$99
Jetson AGX Xavier Developer Kit
Jetson AGX Xavier Developer Kit

Jetson Nano vs Jetson TX1, TX2 and AGX Xavier

FeaturesJetson NanoJetson TX1Jetson TX2 / TX2iJetson AGX Xavier
CPUARM Cortex-A57 (quad-core) @ 1.43GHzARM Cortex-A57 (quad-core) @ 1.73GHzARM Cortex-A57 (quad-core) @ 2GHz +NVIDIA Denver2 (dual-core) @ 2GHzNVIDIA Carmel ARMv8.2 (octal-core) @ 2.26GHz(4x2MB L2 + 4MB L3)
GPU128-core NVIDIA Maxwell @ 921MHz256-core NVIDIA Maxwell @ 998MHz256-core NVIDIA Pascal @ 1300MHz512-core Volta @ 1377 MHz + 64 Tensor Cores
DLNVIDIA GPU support (CUDA, cuDNN, TensorRT)NVIDIA GPU support (CUDA, cuDNN, TensorRT)NVIDIA GPU support (CUDA, cuDNN, TensorRT)Dual NVIDIA Deep Learning Accelerators
Memory4GB 64-bit LPDDR4 @ 1600MHz | 25.6 GB/s4GB 64-bit LPDDR4 @ 1600MHz | 25.6 GB/s8GB 128-bit LPDDR4 @ 1866Mhz | 58.3 GB/s32GB 256-bit LPDDR4x @ 2133MHz | 137GB/s
StorageMicroSD card16GB eMMC 5.132GB eMMC 5.132GB eMMC 5.1
VisionNVIDIA GPU support (CUDA, VisionWorks, OpenCV)NVIDIA GPU support (CUDA, VisionWorks, OpenCV)NVIDIA GPU support (CUDA, VisionWorks, OpenCV)7-way VLIW Vision Accelerator
Encoder(1x) 4Kp30, (2x) 1080p60, (4x) 1080p30(1x) 4Kp30, (2x) 1080p60, (4x) 1080p304Kp60, (3x) 4Kp30, (4x) 1080p60, (8x) 1080p30(4x) 4Kp60, (8x) 4Kp30, (32x) 1080p30
Decoder4Kp60, (2x) 4Kp30, (4x) 1080p60, (8x) 1080p304Kp60, (2x) 4Kp30, (4x) 1080p60, (8x) 1080p30(2x) 4Kp60, (4x) 4Kp30, (7x) 1080p60(2x) 8Kp30, (6x) 4Kp60, (12x) 4Kp30
Camera12 lanes MIPI CSI-2 | 1.5 Gbps per lane12 lanes MIPI CSI-2 | 1.5 Gbps per lane12 lanes MIPI CSI-2 | 2.5 Gbps per lane16 lanes MIPI CSI-2 | 6.8125Gbps per lane
Display2x HDMI 2.0 / DP 1.2 / eDP 1.2 | 2x MIPI DSI2x HDMI 2.0 / DP 1.2 / eDP 1.2 | 2x MIPI DSI2x HDMI 2.0 / DP 1.2 / eDP 1.2 | 2x MIPI DSI(3x) eDP 1.4 / DP 1.2 / HDMI 2.0 @ 4Kp60
WirelessM.2 Key-E site on carrier802.11a/b/g/n/ac 2×2 867Mbps | Bluetooth 4.0802.11a/b/g/n/ac 2×2 867Mbps | Bluetooth 4.1M.2 Key-E site on carrier
Ethernet10/100/1000 BASE-T Ethernet10/100/1000 BASE-T Ethernet10/100/1000 BASE-T Ethernet10/100/1000 BASE-T Ethernet
USB(4x) USB 3.0 + Micro-USB 2.0USB 3.0 + USB 2.0USB 3.0 + USB 2.0(3x) USB 3.1 + (4x) USB 2.0
PCIePCIe Gen 2 x1/x2/x4PCIe Gen 2 x5 | 1×4 + 1×1PCIe Gen 2 x5 | 1×4 + 1×1 or 2×1 + 1×2PCIe Gen 4 x16 | 1×8 + 1×4 + 1×2 + 2×1
CANNANADual CAN bus controllerDual CAN bus controller
Misc IOUART, SPI, I2C, I2S, GPIOsUART, SPI, I2C, I2S, GPIOsUART, SPI, I2C, I2S, GPIOsUART, SPI, I2C, I2S, GPIOs
Socket260-pin edge connector, 45x70mm400-pin board-to-board connector, 50x87mm400-pin board-to-board connector, 50x87mm699-pin board-to-board connector, 100x87mm
Thermals-25°C to 80°C-25°C to 80°C-25°C to 80°C-25°C to 80°C
Power5/10W10W7.5W10/15/30W
Perf472 GFLOPS1 TFLOPS1.3 TFLOPS32 TeraOPS
Jetson AGX Xavier Developer Kit
Jetson AGX Xavier Developer Kit

Jetson AGX Xavier Developer Kit Benchmarks

Phoronix Test Suite :

So I Started with Some of the Standard Tests of the Phoronix Test Suite. You can also run the same test by executing the below commands.

sudo apt-get install -y php-cli php-xml

# Download PTS  https://www.phoronix-test-suite.com/ and  Install

phoronix-test-suite benchmark 1809111-RA-ARMLINUX005

# Accept for Dependency Installation and Wait a few hours, Test Result is available at : 

~/.phoronix-test-suite/test-results/
TestTX1 AGX XavierNotes
Tinymembench (memcpy)3862.06103.9MB/s, More Is Better
Tinymembench (memset)10374.012046.4MB/s, More Is Better
TTSIOD 3D Renderer45.0573.29FPS, More Is Better
7-Zip Compression452610219MIPS, More Is Better
C-Ray851328.39Seconds, Lower Is Better
Primesieve401158.96Seconds, Lower Is Better
AOBench165161.61Seconds, Lower Is Better
Zstd Compression441823.43Seconds, Lower Is Better
FLAC Audio Encoding78.8690Seconds, Lower Is Better
LAME MP3 Encoding113.1460.99Seconds, Lower Is Better
Perl (Pod2html)0.60070.5707Seconds, Lower Is Better
Perl (Interpreter) 0.002358690.00769436Seconds, Lower Is Better
PostgreSQL (Read Only)16079.015775.4Seconds, Lower Is Better
PostgreSQL (Read Write)1222.01248.17Seconds, Lower Is Better
Redis (GET)484688680922Seconds, Lower Is Better
Redis (SET) 333535497367Seconds, Lower Is Better
PyBench63486359Milliseconds, Lower Is Better
Hierarchical INTegration88381313.00120335253.36QUIPs, More Is Better
PHPBench11197992165Score, More Is Better
Git23.6622.52Seconds, Lower Is Better
Scikit-Learn434474Seconds, Lower Is Better
Jetson AGX Xavier Developer Kit
Jetson AGX Xavier Developer Kit

Categories: BenchmarkNvidia

Crazy Engineer

MAKER - ENGINEER - YOUTUBER

0 Comments

Leave a Reply

Avatar placeholder

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.