New COM Express Modules Target Edge AI and Embedded Systems
Adlink has unveiled its latest COM Express modules powered by Intel Core Ultra 9 processors, designed for edge AI, robotics, and industrial automation applications. These boards deliver high graphics throughput, AI processing, and extended temperature resilience, making them suitable for demanding embedded environments.
Express-PTL Family: High-Performance Hybrid CPUs for Industrial Control
The Express-PTL series features Intel’s Core Ultra Series 3 “Panther Lake H” processors, supporting up to 16 CPU cores. Its hybrid design combines four performance cores, eight efficiency cores, and four low-power cores. This architecture balances peak performance and sustained operation under thermal and power constraints, making it ideal for industrial automation and factory control systems.
The module supports up to 128 GB DDR5 SO-DIMM memory with in-band error correction (IBECC), ensuring reliability under continuous load. Industrial variants operate between -40°C and 85°C, targeting outdoor equipment, autonomous machines, and systems exposed to vibration or shock.
Integrated GPU and NPU Accelerate AI Workloads
Adlink integrates Intel Xe3 GPU and NPU 5.0 engines for AI and graphics-intensive tasks. The NPU delivers up to 50 TOPS for AI inference, while the Xe3 GPU achieves up to 120 TOPS with 12 graphics engines. This combination supports computer vision, robotics, medical imaging, visual inspection, and infotainment systems, enabling real-time decision-making at the edge.
Moreover, PCIe connectivity allows low-latency interfacing with external accelerators or motion-control devices. Industrial automation and robotics applications benefit from faster AI inference and seamless integration with control systems.
Industrial-Grade Features for Reliability and Deterministic Control
Adlink targets time-sensitive industrial automation with features like Time-Sensitive Networking (TSN), functional safety frameworks, and extended-temperature components. These capabilities ensure deterministic performance for factory automation, robotics, and smart transport.
The ruggedized design reduces failure risk in extreme conditions, while integrated diagnostic support and remote monitoring simplify predictive maintenance. This approach aligns with modern PLC and DCS standards, improving reliability and lifecycle efficiency.
COM-HPC Mini: Compact AI Module for Space-Constrained Designs
Adlink plans a follow-on COM-HPC Mini module, measuring 95 mm by 70 mm. The COM-HPC-mPTL integrates CPU, GPU, NPU, and soldered memory, ideal for autonomous vehicles, defence platforms, and other vibration-prone or temperature-sensitive environments.
This module targets compact AI deployments in industrial automation, robotics, and transportation. Its small form factor does not compromise performance, supporting graphics-heavy and AI-intensive applications where space and thermal management are critical.
Development Kits Enable Rapid Prototyping
Adlink will release development kits for both Express-PTL and COM-HPC-mPTL modules, including reference carrier boards and full I/O support. These kits accelerate prototyping and integration for edge AI, embedded control, and factory automation systems.
By aligning module releases with Intel’s Panther Lake embedded platforms, Adlink positions these boards for immediate adoption in industrial automation, robotics, and smart transportation applications.
Author Insights: Trends in Edge AI and Industrial Embedded Systems
The release of Intel Core Ultra-powered modules signals a shift toward AI-enabled industrial automation. Edge computing now requires hybrid CPU cores, high-throughput GPUs, and dedicated NPUs to handle real-time analytics in harsh environments.
From experience in industrial control system deployments, integrating AI processing on compact, rugged modules significantly reduces latency and hardware complexity. Adlink’s approach supports both standard embedded computing and AI-specific applications, offering flexibility for control engineers and automation integrators.
Application Scenarios
-
Robotics: Real-time navigation, object recognition, and motion control for humanoid and quadruped robots.
-
Industrial Automation: Factory automation, PLC/DCS integration, and predictive maintenance systems.
-
Smart Transport: Edge AI in autonomous vehicles, traffic monitoring, and visual inspection platforms.