ADLINK to Showcase Scalable Edge AI Computing at Embedded World 2026

Edge AI computing platforms for industrial automation and robotics
In industrial automation, robotics, and autonomous systems, edge computing platforms must process AI workloads close to the data source while maintaining reliability and scalability. ADLINK Technology will present its latest portfolio of edge AI computing platforms at Embedded World 2026 (10–12 March, Nuremberg, Germany, Hall 3, Booth 147).

The showcase highlights a unified edge AI architecture designed to support deployment across a wide range of hardware form factors, from compact computer-on-module (COM) systems to workstation-class GPU servers. The portfolio aims to address the increasing demand for computing platforms capable of supporting real-time AI inference in industrial environments.

Embedded computing platforms based on Intel processors
A key focus of the portfolio is a series of embedded computing platforms built on Intel Core Ultra Series 3 processors, which integrate neural processing unit (NPU) acceleration for edge AI workloads.

ADLINK’s embedded modules include Express-PTL and COM-HPC-mPTL platforms, which provide a compact computing foundation with DDR5 memory support for embedded and edge AI applications. These modules are designed for integration into industrial machines, robotics systems, and edge computing devices requiring low-latency AI processing.

The portfolio also includes SBC35-PTL 3.5-inch single-board computers and VNX-PTL processor blades, which extend computing performance for rugged industrial environments where reliable edge processing is required.

Embedded systems for industrial visualization and real-time inference
Beyond modular hardware, ADLINK also presents complete embedded systems designed for real-time AI inference and industrial automation tasks.

Examples include the MXE-330 compact edge AI computer, which is designed for space-constrained industrial installations. Panel PC platforms such as SP2-TWL and STC2-TWL, based on Intel Twin Lake processors, support applications including industrial visualization, machine control interfaces, and AI-enabled monitoring systems.

Such platforms enable edge computing capabilities directly within production environments, reducing reliance on centralized cloud processing and supporting real-time decision-making.

High-performance edge servers for AI acceleration
For more demanding AI workloads, ADLINK will present computing platforms based on Intel Xeon 600 processors for workstations. These platforms support multi-GPU configurations for high-performance AI inference and large-scale edge computing deployments.

The portfolio includes the ISB-W890 server motherboard and the AXE GPU server series, which are designed to deliver high memory bandwidth and scalable GPU acceleration. Such systems are intended for applications requiring high-density AI computing, including industrial inspection, autonomous systems, and data-intensive analytics.

Robotics and autonomous AI platforms with NVIDIA technology
In addition to Intel-based systems, ADLINK will demonstrate robotics computing platforms powered by NVIDIA Jetson Thor processors.

The DLAP-700 Series, based on Jetson Thor architecture, supports real-time perception and AI processing for robotics applications. These platforms are designed for humanoid robots, collaborative industrial robots, and other autonomous systems that require rapid sensor processing and adaptive AI capabilities.

Such computing platforms enable robots to analyze visual and sensor data directly at the edge, allowing faster response times compared with cloud-based processing approaches.

Edge computing for UAV and aerial systems
ADLINK will also present a UAV platform developed in collaboration with Xdynamics. The system integrates flight control, RF communication, and edge computing capabilities within a compact hardware platform.

This architecture demonstrates how scalable edge AI computing can support applications beyond ground-based robotics, extending to aerial systems that require onboard processing for navigation, perception, and autonomous control.

Preparing for next-generation GPU acceleration
Looking ahead, ADLINK will preview future hardware developments, including MXM graphics modules based on NVIDIA Blackwell architecture. These platforms are designed to support increasing AI computing density requirements as next-generation GPU technologies become available.

By integrating embedded modules, industrial edge systems, and high-performance GPU servers within a unified architecture, ADLINK aims to provide scalable computing solutions for AI deployments ranging from compact embedded devices to large-scale edge computing infrastructures.

www.adlinktech.com