Powering AI in Edge Applications
Deep Learning Acceleration Platforms with NVIDIA® Jetson™ SOMs
ADLINK, as the pioneer in the edge computing, AI-focused industry leader, and a partner to NVIDIA, is committed to deliver the edge AI solutions of heterogeneous computing architecture and hardware acceleration for deep learning, inference, machine learning, automated optical inspection, and other decision-making workloads.
ADLINK’s products are designed for durability in harsh industrial and embedded environments, operating at extended temperature, vibration, and relative-humidity ranges, enabling systems integrators as well as OEMs/ODMs to lower development costs and shorten the time to market.
Why ADLINK
With 26-year dedication in automation, ADLINK realizes the demands and challenges customer face, and has established alliances with mega manufacturers like AUO and Foxconn to deliver proven and optimized solutions.
ADLINK partners with AI system integrators and software/cloud service providers to expand the service scope for customized AI applications, centralized device management, and cloud connectivity.
ADLINK Edge AI Platforms feature high-performance yet energy-efficient in compact, fanless, and durable packages. Topped with the no-moving-parts design and wide-temp., they provide 24/7 reliable operations.
WHAT ADLINK Can Do for You
Deep-Learning Consultancy Services
ADLINK offers consultancy services via deep learning profiling to help users determine the most suitable hardware platform to satisfy their performance and size, weight, and power (SWaP) requirements. Considering the types of neural networks, batch size and other inputs, ADLINK’s profiling tool models the performance of computing platforms. The tool runs these inputs against a large database of AI and deep learning accelerators, and generates statistics, such as inferences per second; performance per watt; and performance per dollar, for popular neural networks.
-
Off-the-Shelf and ODM/OEM
-
Off-the-Shelf and ODM/OEM
The ADLINK Jetson-based computers are compact, ruggedized, thermally-optimized, and supported with extended lifecycles, ideal to run AI inferencing, facial recognition, object detection, or other AI-based applications at the edge.
-
Customization is also available for different vertical markets, enabling customers to rapidly harness the power of AI at the edge. With long-term success in designing embedded modules, carrier boards and systems, ADLINK caters individual projects by quickly developing edge AI platforms based on the latest NVIDIA Jetson technologies.
-
ROS 2 Robotics Solutions
-
ROS 2 Robotics Solutions
ADLINK provide lines of ROS 2 enabled Robotic Controllers, supporting the full complement of resources developed with the NVIDIA JetPack SDK and ADLINK’s Neuron SDK, and are specifically suited for robotic applications demanding high-AI computing with minimal power consumption.
-
Jetson Developer Kits Available
-
Jetson Developer Kits Available
ADLINK’s Jetson developer kits help users design, test and deploy hardware/software applications for multiple use scenarios. Each Jetson developer kit includes a pin-to-pin NVIDIA Jetson SoM with heat sink/fan for cooling, a durable carrier board with lines of industrial grade I/O ports and expansion slots, and an adapter with power cord. It also comes along with Linux Ubuntu operating system, NVIDIA Jetpack SDK as well as Allxon device management software.
NVIDIA® Jetson Edge AI Platforms We Offer
The Jetson-based DLAP Edge AI Platforms offer users the flexibility to cost-effectively achieve the SWaP and AI performance. Users can choose from lines of NVIDIA Jetson SoMs, form factors, and power consumption ranges to satisfy their specific application requirements.
-
DLAP-211-Nano
Edge AI Platform Powered by NVIDIA® Jetson Nano™- 128-core Maxwell™ GPU
- 4-core ARM® Cortex-A57 CPU
- Compact, robust and fanless design
- Wide temperature operation support from -20°C to 70°C
- Integrated with Allxon remote device management
DLAP-301-Nano
AI-enabled Embedded NVR Powered by NVIDIA® Jetson Nano™- 128-core Maxwell™ GPU
- 4-core ARM® Cortex-A57 CPU
- 8 ports of PoE for video streams
- 2.5” SATA SSD slot
- Industrial I/O ports and visual inferencing capabilities
ROScube Pico Nano
NVIDIA® Jetson Nano™ SOM-based platform for rapid development of ROS and AI applications- 128-core Maxwell™ GPU
- 4-core ARM® Cortex-A57 CPU
- Compact, robust and fanless design
- Designed for rapid development and deployment of ROS and AI applications
- Full complement of resources provided by ADLINK’s Neuron SDK, Neuron IDE, and Neuron Library
-
DLAP-201-JT2
Edge AI Platform Powered by NVIDIA® Jetson™ TX2- 256-core Pascal™ GPU with 256 CUDA Cores
- Dual-core NVIDIA Denver 2 64-Bit CPU, Quad-Core ARM® Cortex®-A57 MPCore
- Compact, robust and fanless design
- Wide-temp operation support from -20°C to 70°C
- Integrated with Allxon remote device management
-
DLAP-211-JNX
Edge AI Platform Powered by NVIDIA® Jetson Xavier™ NX- 384-core Volta™ GPU with 48 Tensor Cores
- 6-core ARM® v8.2 64-bit CPU
- Compact, robust and fanless design
- Wide-temp operation support from -20°C to 70°C
- AWS IoT Greengrass service and Allxon remote device management available
DLAP-301-JNX
AI-enabled Embedded NVR Powered by NVIDIA® Jetson Xavier™ NX- 384-core Volta™ GPU with 48 Tensor Cores
- 6-core ARM® v8.2 64-bit CPU
- 8 ports of PoE for video streams
- 2.5” SATA SSD slot
- Industrial I/O ports and visual inferencing capabilities
- Amazon Lookout for Vision and Allxon remote device management available
ROScube Pico NX
NVIDIA® Jetson Xavier™ NX SOM-based platform for rapid development of ROS and AI applications- 384-core Volta™ GPU with 48 Tensor Cores
- 6-core ARM® v8.2 64-bit CPU
- Compact, robust and fanless design
- Designed for rapid development and deployment of ROS and AI applications
- Full complement of resources provided by ADLINK’s Neuron SDK, Neuron IDE, and Neuron Library
-
DLAP-401-Xavier
Edge AI Platform Powered by NVIDIA® Jetson AGX Xavier™- 512-core Volta™ GPU with 64 Tensor Cores
- 8-core ARM® v8.2 64-bit CPU
- Compact, robust and fanless design
- Wide-temp operation support from -20°C to 60°C
- Integrated with Allxon for remote device management
ROScube-X
Embedded Robotic Controller Powered by NVIDIA® Jetson AGX Xavier™ Module- 512-core Volta™ GPU with 64 Tensor Cores
- 8-core ARM® v8.2 64-bit CPU
- Designed for rapid development and deployment of ROS and AI applications
Resources
Application Story
Artificial Intelligence and Computer Vision Speeds Canteen/Restaurant Checkout
Download
Video
Discover ADLINK’s deep learning acceleration platforms (DLAP) in the one-minute video.
Watch Now
Video
Watch the application story of how AI and computer vision speeds restaurant checkout.
Watch Now
