IoT Edge AI Computers from ASUS Enhance Generative AI Performance with NVIDIA Jetson Orin
Super mode on the NVIDIA Jetson Orin NX and Orin Nano series allows you to experience 2X greater Gen AI model performance.
- The ASUS IoT edge AI computer models PE1102N, PE1101N, and PE1100N support Super mode for NVIDIA Jetson Orin NX and Orin Nano series, which results in up to 2X Gen AI model performance.
- For devices with NVIDIA Jetson Orin NX 16GB modules, super mode offers up to 157 TOPS (Sparse).
- The most recent NVIDIA JetPack 6.2 SDK is now supported by all ASUS IoT edge AI computers driven by Jetson Orin.
Leading AIoT solutions provider ASUS IoT revealed that its edge AI computers, which are powered by the NVIDIA Jetson Orin platform, have been updated to support the new Super mode. This update boosts Gen AI model performance by up to 2X, memory bandwidth by up to 1.5X, and AI TOPS by up to 1.7X.
Significant compatibility with Super mode and up to 1.7X AI TOPS
A variety of Gen AI models, including as large language models (LLMs), visual language models (VLMs), and vision transformers (ViTs), are made possible by Super mode, which increases AI TOPS by up to 1.7X. Popular machine learning frameworks like Hugging Face Transformers, Ollama, llama.cpp, vLLM, MLC, NVIDIA TensorRT-LLM, and others are supported by Jetson Orin modules. They are therefore great choices for robotics, multimodal systems, and edge applications.
NVIDIA Jetson-powered ASUS IoT edge AI computers are faster, more adaptable, and more durable
The ASUS IoT NVIDIA Jetson edge AI models PE1102N, PE1101N, and PE1100N can now process data more efficiently and tackle Gen AI tasks at much higher rates. These models are perfect for smart city, smart transportation, and in-vehicle and smart logistics applications including smart parking, traffic analysis, and AGV/AMR because of their small size, high degree of adaptability, and wide connection. Developers may handle data at the edge more quickly, with more security, and with considerable total cost of ownership (TCO) reductions with each solution.
In addition to their robust architecture and ability to function in a broad variety of temperatures, ASUS IoT edge AI computers powered by the NVIDIA Jetson series offer exceptional thermal performance to optimise overall efficiency for outdoor use.
Super mode is enabled by supporting the most recent NVIDIA JetPack 6.2 SDK
By downloading the most recent NVIDIA JetPack 6.2 SDK, ASUS IoT edge AI computers (PE1102N, PE1101N, and PE1100N) may enable super mode for the NVIDIA Jetson Orin NX and Orin Nano series.
The most recent production release of JetPack 6 is JetPack 6.2. With a root file system based on Ubuntu 22.04 and the Linux Kernel 5.15, Jetson Linux 36.4.3 is part of this version. CUDA 12.6, NVIDIA TensorRT 10.3, NVIDIA cuDNN 9.3, NVIDIA Vision Programming Interface 3.2, NVIDIA Deep Learning Accelerator 3.1, and DLFW 24.0 are all part of the Jetson AI stack that comes with JetPack 6.2. The NVIDIA Jetson Orin Nano and NVIDIA Jetson Orin NX production modules now have new high-power Super modes with JetPack 6.2. Super modes can increase the performance of Gen AI models’ AI inference by up to two times.
PE1102N, PE1101N and PE1100N Specifications
PE1102N | PE1101N | PE1100N | ||
Case | Dimension | 152 x 114 x 72mm | 130 x 90 x 72mm | 152 x 114 x 72mm |
System | Processor | NVIDIA Jetson Orin Nano NVIDIA Jetson Orin NX | NVIDIA Jetson Orin Nano NVIDIA Jetson Orin NX | NVIDIA Jetson Orin Nano NVIDIA Jetson Orin NX |
Chipset | 6/8-core Arm Cortex -A78AE | 6/8-core Arm Cortex -A78AE | 6/8-core Arm Cortex -A78AE | |
Graphics | NVIDIA Ampere GPU with Tensor Cores | NVIDIA Ampere GPU with Tensor Cores | NVIDIA Ampere GPU with Tensor Cores | |
Memory | on-board, up to 16GB LPDDR5 | on-board, up to 16GB LPDDR5 | on-board, up to 16GB LPDDR5 | |
I/O Interface | PoE | – | – | – |
Ethernet | 2 x 10/100/1000Mbps, RJ45 | 1 x 10/100/1000Mbps, RJ45 | 2 x 10/100/1000Mbps, RJ45 | |
Display Port | 1x HDMI | 1 x HDMI | 1x HDMI | |
Serial Port | 1 x DB25 : RS-232/422/485 & CAN bus | 1 x DB9: CAN bus | 2 x DB9: RS-232/422/485 1 x DB9: CAN bus | |
USB 2.0 | 1 x USB 2.0, Micro-USB for OS Flash 2 x USB 2.0, Pin Header (Internal) | 1 x USB 2.0, Type-C for OS Flash | 1 x USB 2.0, Micro-USB for OS Flash 2 x USB 2.0, Pin Header (Internal) | |
USB 3.2/ 3.1 | 3 x USB 3.2 Gen1 (5Gbps), Type-A | 2 x USB 3.2 Gen2 (10Gbps), Type-A | 3 x USB 3.2 Gen1 (5Gbps), Type-A | |
Audio | – | – | – | |
Digital I/O | 4 x DI, 4 x DO (2×5 Terminal Block, w/ isolation) | – | 4 x DI, 4 x DO (2×5 Terminal Block, w/ isolation) | |
GPIO | – | 1 x DB9: GPIO/I2C/UART | – | |
Storage Interface | SATA HDD | – | – | – |
mSATA | – | – | – | |
M.2 (M-key) | 1 (NVMe) | 1 (NVMe) | 1 (NVMe) | |
eMMC | – | – | – | |
SD Card | – | – | – | |
Expansion | mPCIe | – | – | – |
M.2 | 1 x M.2 E-key, 1 x M.2 B-key | 1 x M.2 E-key | 1 x M.2 E-key, 1 x M.2 B-key | |
SIM | 2 | – | 2 | |
PCI/PCIe | – | – | – | |
MXM | – | – | – | |
Others | 1 x AEM (GSML or OOB or LAN) | – | 1 x AEM (LAN) | |
Power Supply | DC Input | 12-36V DC | 12-24V DC | 12-24V DC |
Ignition Control | Integrated | – | – | |
Environmental | Operating Temp. | -25–60°C | -25–55°C | -25–60°C |
Certification | CE, FCC, CB, BSMI | CE, FCC, CB, BSMI | CE, FCC, CB, BSMI | |
Shock & Vibration | MIL-STD 810H, and 5-500Hz; 5 Grms | MIL-STD 810H, and 5-500Hz; 5 Grms | MIL-STD 810H, and 5-500Hz; 5 Grms |