ADLINK has introduced its Express-PTL COM Express modules, powered by Intel’s Core Ultra Series 3 processors and featuring an integrated Intel Arc GPU for high-performance edge computing. The new platform is designed to handle complex AI workloads with up to 180 TOPS of combined compute, including 50 TOPS from the NPU 5.0 and 120 TOPS from the GPU, while maintaining industrial-grade reliability in demanding environments.
The Express-PTL modules leverage a hybrid CPU architecture with four high-performance (P) cores, eight efficiency (E) cores, and four low-power (LPE) cores, optimizing power consumption and thermal efficiency. They support up to 128 GB of DDR5 SO-DIMM memory with In-Band ECC for low-latency data processing, making them suitable for applications requiring both high-speed computation and robust error correction.
Industrial ruggedness is a key feature, with the modules supporting operating temperatures from -40°C to 85°C and compliance with TCC, TSN, and FuSa/FSEDP standards. This makes them ideal for autonomous robots, industrial automation, and outdoor edge deployments where environmental resilience is critical.
Beyond raw performance, the Express-PTL modules simplify system design by integrating graphics acceleration, reducing the need for external GPUs in applications like medical imaging or infotainment. The PCIe interface also enables low-latency accelerator interactions, which is essential for real-time decision-making in robotics and autonomous systems.
ADLINK’s lineup will expand with additional platforms later this year, including the COM-HPC-mPTL (launching Q2 2026) and SBC35-PTL (Q3 2026), both featuring soldered-down memory for enhanced durability. The MXE-330, a versatile edge computing platform with customizable I/O, is also set to debut in the third quarter.
Development kits, including reference carriers and full I/O support, will be available starting Q2 2026, allowing engineers to accelerate prototyping for AI-driven edge applications.
- Processor: Intel Core Ultra Series 3 (hybrid architecture: 4 P-cores, 8 E-cores, 4 LPE cores)
- GPU: Intel Arc GPU (up to 12 Xe cores, 120 TOPS AI performance)
- NPU: NPU 5.0 (50 TOPS dedicated AI acceleration)
- Total AI Compute: Up to 180 platform TOPS
- Memory: Up to 128 GB DDR5 SO-DIMM with IBECC and In-Band ECC
- Industrial Features: -40°C to 85°C temperature range, TCC/TSN support, FuSa/FSEDP compliance
- Connectivity: PCIe interface for low-latency accelerator interaction
The Express-PTL modules strike a balance between performance and efficiency, making them well-suited for edge AI workloads that demand both computational power and environmental resilience. The inclusion of an integrated Arc GPU also reduces system complexity and cost, which is particularly beneficial for applications where space or budget constraints are tight.
This platform will be most valuable in industries such as robotics (including humanoid and quadruped robots), industrial automation, and outdoor edge computing, where real-time processing and rugged operation are non-negotiable. For developers working on AI-driven solutions, the Express-PTL’s combination of high TOPS, DDR5 memory support, and industrial-grade durability positions it as a strong contender in the next wave of embedded AI hardware.
Availability for the Express-PTL modules is not yet confirmed, but development kits are expected to ship in Q2 2026. Pricing details will be announced closer to launch.
