ARMv7 rev 3 testing with a ODROID-XU4 Hardkernel Odroid XU4 and llvmpipe 2GB on Ubuntu 18.04 via the Phoronix Test Suite.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 1903179-HV-1903178SP62
Jetson Prep
ARMv7 rev 3 testing with a ODROID-XU4 Hardkernel Odroid XU4 and llvmpipe 2GB on Ubuntu 18.04 via the Phoronix Test Suite.
Jetson AGX Xavier:
Processor: ARMv8 rev 0 @ 2.27GHz (8 Cores), Motherboard: jetson-xavier, Memory: 16384MB, Disk: 31GB HBG4a2, Graphics: NVIDIA Tegra Xavier, Monitor: VE228
OS: Ubuntu 18.04, Kernel: 4.9.108-tegra (aarch64), Desktop: Unity 7.5.0, Display Server: X Server 1.19.6, Display Driver: NVIDIA 31.0.2, OpenGL: 4.6.0, Vulkan: 1.1.76, Compiler: GCC 7.3.0 + CUDA 10.0, File-System: ext4, Screen Resolution: 1920x1080
Jetson TX2 Max-P:
Processor: ARMv8 rev 3 @ 2.04GHz (4 Cores / 6 Threads), Motherboard: quill, Memory: 8192MB, Disk: 31GB 032G34, Graphics: NVIDIA TEGRA, Monitor: VE228
OS: Ubuntu 16.04, Kernel: 4.4.38-tegra (aarch64), Desktop: Unity 7.4.0, Display Server: X Server 1.18.4, Display Driver: NVIDIA 28.2.1, OpenGL: 4.5.0, Compiler: GCC 5.4.0 20160609 + CUDA 9.0, File-System: ext4, Screen Resolution: 1920x1080
Jetson TX2 Max-Q:
Processor: ARMv8 rev 3 @ 1.27GHz (4 Cores / 6 Threads), Motherboard: quill, Memory: 8192MB, Disk: 31GB 032G34, Graphics: NVIDIA TEGRA, Monitor: VE228
OS: Ubuntu 16.04, Kernel: 4.4.38-tegra (aarch64), Desktop: Unity 7.4.0, Display Server: X Server 1.18.4, Display Driver: NVIDIA 28.2.1, OpenGL: 4.5.0, Compiler: GCC 5.4.0 20160609 + CUDA 9.0, File-System: ext4, Screen Resolution: 1920x1080
Raspberry Pi 3 Model B+:
Processor: ARMv7 rev 4 @ 1.40GHz (4 Cores), Motherboard: BCM2835 Raspberry Pi 3 Model B Plus Rev 1.3, Memory: 926MB, Disk: 32GB GB2MW, Graphics: BCM2708
OS: Raspbian 9.6, Kernel: 4.19.23-v7+ (armv7l), Desktop: LXDE, Display Server: X Server 1.19.2, Compiler: GCC 6.3.0 20170516, File-System: ext4, Screen Resolution: 656x416
ASUS TinkerBoard:
Processor: ARMv7 rev 1 @ 1.80GHz (4 Cores), Motherboard: Rockchip (Device Tree), Memory: 2048MB, Disk: 32GB GB1QT
OS: Debian 9.0, Kernel: 4.4.16-00006-g4431f98-dirty (armv7l), Desktop: LXDE, Display Server: X Server 1.18.4, Compiler: GCC 6.3.0 20170516, File-System: ext4, Screen Resolution: 1024x768
Jetson TX1 Max-P:
Processor: ARMv8 rev 1 @ 1.73GHz (4 Cores), Motherboard: jetson_tx1, Memory: 4096MB, Disk: 16GB 016G32, Graphics: NVIDIA Tegra X1, Monitor: VE228
OS: Ubuntu 16.04, Kernel: 4.4.38-tegra (aarch64), Desktop: Unity 7.4.5, Display Server: X Server 1.18.4, Display Driver: NVIDIA 28.1.0, OpenGL: 4.5.0, Vulkan: 1.0.8, Compiler: GCC 5.4.0 20160609, File-System: ext4, Screen Resolution: 1920x1080
ODROID-XU4:
Processor: ARMv7 rev 3 @ 1.50GHz (8 Cores), Motherboard: ODROID-XU4 Hardkernel Odroid XU4, Memory: 2048MB, Disk: 16GB AJTD4R, Graphics: llvmpipe 2GB, Monitor: VE228
OS: Ubuntu 18.04, Kernel: 4.14.37-135 (armv7l), Display Server: X Server 1.19.6, OpenGL: 3.3 Mesa 18.0.0-rc5 (LLVM 6.0 128 bits), Compiler: GCC 7.3.0, File-System: ext4, Screen Resolution: 1920x1080
C-Ray 1.1
Total Time - 4K, 16 Rays Per Pixel
Seconds < Lower Is Better
Jetson AGX Xavier ....... 355 |========
Jetson TX2 Max-P ........ 585 |==============
Jetson TX2 Max-Q ........ 869 |====================
Raspberry Pi 3 Model B+ . 2030 |===============================================
ASUS TinkerBoard ........ 1718 |========================================
Jetson TX1 Max-P ........ 753 |=================
ODROID-XU4 .............. 827 |===================
LeelaChessZero 0.20.1
Backend: BLAS
Nodes Per Second > Higher Is Better
Jetson AGX Xavier . 47.62 |====================================================
Rust Prime Benchmark
Prime Number Test To 200,000,000
Seconds < Lower Is Better
Jetson AGX Xavier ....... 32.37 |=
Jetson TX2 Max-P ........ 104.96 |===
Jetson TX2 Max-Q ........ 170.25 |====
Raspberry Pi 3 Model B+ . 1097.69 |===========================
ASUS TinkerBoard ........ 1821.05 |============================================
Jetson TX1 Max-P ........ 128.45 |===
ODROID-XU4 .............. 574.11 |==============
TTSIOD 3D Renderer 2.3b
Phong Rendering With Soft-Shadow Mapping
FPS > Higher Is Better
Jetson AGX Xavier ....... 133.00 |=============================================
Jetson TX2 Max-P ........ 49.26 |=================
Jetson TX2 Max-Q ........ 28.85 |==========
Raspberry Pi 3 Model B+ . 17.66 |======
ASUS TinkerBoard ........ 21.22 |=======
Jetson TX1 Max-P ........ 45.09 |===============
ODROID-XU4 .............. 41.96 |==============
CUDA Mini-Nbody 2015-11-10
Test: Original
(NBody^2)/s > Higher Is Better
Jetson AGX Xavier . 47.13 |====================================================
Jetson TX2 Max-P .. 8.24 |=========
Jetson TX2 Max-Q .. 6.77 |=======
NVIDIA TensorRT Inference
Neural Network: ResNet152 - Precision: INT8 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 493.22 |===================================================
Jetson TX2 Max-P .. 22.07 |==
Jetson TX2 Max-Q .. 17.36 |==
NVIDIA TensorRT Inference
Neural Network: VGG19 - Precision: INT8 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 394.66 |===================================================
Jetson TX2 Max-P .. 15.92 |==
Jetson TX2 Max-Q .. 12.59 |==
PyBench 2018-02-16
Total For Average Test Times
Milliseconds < Lower Is Better
Jetson AGX Xavier ....... 3007 |=======
Jetson TX2 Max-P ........ 5408 |============
Jetson TX2 Max-Q ........ 8735 |===================
Raspberry Pi 3 Model B+ . 20913 |==============================================
ASUS TinkerBoard ........ 11502 |=========================
Jetson TX1 Max-P ........ 6339 |==============
ODROID-XU4 .............. 5009 |===========
NVIDIA TensorRT Inference
Neural Network: VGG19 - Precision: FP16 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 203.96 |===================================================
Jetson TX2 Max-P .. 29.83 |=======
Jetson TX2 Max-Q .. 23.94 |======
OpenCV Benchmark 3.3.0
Seconds < Lower Is Better
Jetson AGX Xavier ....... 128.00 |===========
Jetson TX2 Max-P ........ 296.00 |==========================
Jetson TX2 Max-Q ........ 493.00 |===========================================
Raspberry Pi 3 Model B+ . 2.74 |
ODROID-XU4 .............. 520.70 |=============================================
NVIDIA TensorRT Inference
Neural Network: VGG16 - Precision: INT8 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 475.08 |===================================================
Jetson TX2 Max-P .. 19.91 |==
Jetson TX2 Max-Q .. 15.79 |==
NVIDIA TensorRT Inference
Neural Network: ResNet152 - Precision: FP16 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 259.82 |===================================================
Jetson TX2 Max-P .. 41.91 |========
Jetson TX2 Max-Q .. 32.67 |======
NVIDIA TensorRT Inference
Neural Network: VGG16 - Precision: FP16 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 247.95 |===================================================
Jetson TX2 Max-P .. 36.87 |========
Jetson TX2 Max-Q .. 29.83 |======
Zstd Compression 1.3.4
Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19
Seconds < Lower Is Better
Jetson AGX Xavier ....... 80.06 |=======
Jetson TX2 Max-P ........ 144.97 |=============
Jetson TX2 Max-Q ........ 253.80 |=======================
Raspberry Pi 3 Model B+ . 342.23 |===============================
ASUS TinkerBoard ........ 496.62 |=============================================
Jetson TX1 Max-P ........ 145.80 |=============
FLAC Audio Encoding 1.3.2
WAV To FLAC
Seconds < Lower Is Better
Jetson AGX Xavier ....... 54.47 |=======
Jetson TX2 Max-P ........ 65.07 |=========
Jetson TX2 Max-Q ........ 104.28 |==============
Raspberry Pi 3 Model B+ . 339.53 |=============================================
ASUS TinkerBoard ........ 279.05 |=====================================
Jetson TX1 Max-P ........ 79.20 |==========
ODROID-XU4 .............. 97.03 |=============
NVIDIA TensorRT Inference
Neural Network: ResNet50 - Precision: FP16 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 547.50 |===================================================
Jetson TX2 Max-P .. 92.28 |=========
Jetson TX2 Max-Q .. 72.01 |=======
NVIDIA TensorRT Inference
Neural Network: ResNet152 - Precision: INT8 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 372.73 |===================================================
Jetson TX2 Max-P .. 18.29 |===
Jetson TX2 Max-Q .. 14.50 |==
NVIDIA TensorRT Inference
Neural Network: VGG16 - Precision: INT8 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 303.78 |===================================================
Jetson TX2 Max-P .. 17.56 |===
Jetson TX2 Max-Q .. 14.24 |==
NVIDIA TensorRT Inference
Neural Network: ResNet152 - Precision: FP16 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 224.19 |===================================================
Jetson TX2 Max-P .. 35.11 |========
Jetson TX2 Max-Q .. 27.34 |======
7-Zip Compression 16.02
Compress Speed Test
MIPS > Higher Is Better
Jetson AGX Xavier ....... 19212 |==============================================
Jetson TX2 Max-P ........ 5593 |=============
Jetson TX2 Max-Q ........ 3294 |========
Raspberry Pi 3 Model B+ . 2013 |=====
ASUS TinkerBoard ........ 2836 |=======
Jetson TX1 Max-P ........ 4508 |===========
ODROID-XU4 .............. 4120 |==========
NVIDIA TensorRT Inference
Neural Network: VGG19 - Precision: INT8 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 265.81 |===================================================
Jetson TX2 Max-P .. 14.32 |===
Jetson TX2 Max-Q .. 11.45 |==
NVIDIA TensorRT Inference
Neural Network: ResNet50 - Precision: INT8 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 1215.08 |==================================================
Jetson TX2 Max-P .. 59.69 |==
Jetson TX2 Max-Q .. 47.15 |==
GLmark2
Resolution: 1920 x 1080
Score > Higher Is Better
Jetson AGX Xavier . 2876 |=====================================================
ODROID-XU4 ........ 1 |
Tesseract OCR 4.0.0-beta.1
Time To OCR 7 Images
Seconds < Lower Is Better
Jetson AGX Xavier . 71.94 |====================
ODROID-XU4 ........ 180.66 |===================================================
NVIDIA TensorRT Inference
Neural Network: GoogleNet - Precision: FP16 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 1006 |=====================================================
Jetson TX2 Max-P .. 233 |============
Jetson TX2 Max-Q .. 179 |=========
NVIDIA TensorRT Inference
Neural Network: VGG16 - Precision: FP16 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 208.76 |===================================================
Jetson TX2 Max-P .. 32.64 |========
Jetson TX2 Max-Q .. 25.99 |======
NVIDIA TensorRT Inference
Neural Network: VGG19 - Precision: FP16 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 172.50 |===================================================
Jetson TX2 Max-P .. 26.56 |========
Jetson TX2 Max-Q .. 21.04 |======
NVIDIA TensorRT Inference
Neural Network: GoogleNet - Precision: INT8 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 1693 |=====================================================
Jetson TX2 Max-P .. 130 |====
Jetson TX2 Max-Q .. 104 |===
NVIDIA TensorRT Inference
Neural Network: GoogleNet - Precision: FP16 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 796 |======================================================
Jetson TX2 Max-P .. 197 |=============
Jetson TX2 Max-Q .. 156 |===========
NVIDIA TensorRT Inference
Neural Network: ResNet50 - Precision: FP16 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 636.00 |===================================================
Jetson TX2 Max-P .. 111.00 |=========
Jetson TX2 Max-Q .. 86.08 |=======
NVIDIA TensorRT Inference
Neural Network: AlexNet - Precision: FP16 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 2038 |=====================================================
Jetson TX2 Max-P .. 462 |============
Jetson TX2 Max-Q .. 374 |==========
NVIDIA TensorRT Inference
Neural Network: AlexNet - Precision: FP16 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 1200 |=====================================================
Jetson TX2 Max-P .. 264 |============
Jetson TX2 Max-Q .. 216 |==========
NVIDIA TensorRT Inference
Neural Network: ResNet50 - Precision: INT8 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 902.78 |===================================================
Jetson TX2 Max-P .. 49.97 |===
Jetson TX2 Max-Q .. 39.15 |==
NVIDIA TensorRT Inference
Neural Network: GoogleNet - Precision: INT8 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 1146.00 |==================================================
Jetson TX2 Max-P .. 113.00 |=====
Jetson TX2 Max-Q .. 88.88 |====
NVIDIA TensorRT Inference
Neural Network: AlexNet - Precision: INT8 - Batch Size: 4 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 1143 |=====================================================
Jetson TX2 Max-P .. 184 |=========
Jetson TX2 Max-Q .. 148 |=======
NVIDIA TensorRT Inference
Neural Network: AlexNet - Precision: INT8 - Batch Size: 32 - DLA Cores: Disabled
Images Per Second > Higher Is Better
Jetson AGX Xavier . 3143 |=====================================================
Jetson TX2 Max-P .. 301 |=====
Jetson TX2 Max-Q .. 237 |====
LeelaChessZero 0.20.1
Backend: CUDA + cuDNN
Nodes Per Second > Higher Is Better
Jetson AGX Xavier . 952.89 |===================================================
LeelaChessZero 0.20.1
Backend: CUDA + cuDNN FP16
Nodes Per Second > Higher Is Better
Jetson AGX Xavier . 2515.01 |==================================================