New industrial boards pair Intel 18A processors with integrated graphics and AI acceleration, bringing faster inference and demanding workloads closer to connected edge environments today.
Advantech has integrated Intel processors built on the 18A process into its latest industrial motherboards, including the MIO-5381 and AIMB-234, targeting faster and more capable edge AI computing. The platforms combine CPU, GPU and NPU resources to handle AI workloads closer to where industrial data is generated.
The new processors are positioned for applications requiring real-time processing without relying entirely on remote computing infrastructure. Advantech says the MIO-5381 can deliver up to 100 platform TOPS for AI acceleration, while its CPU and NPU capabilities support real-time AI inference for workloads such as endoscopic video streams. The approach is aimed at bringing advanced computing into demanding edge environments.
Graphics performance also receives an upgrade. The processors can provide up to 12 Xe cores, enabling up to 50% faster graphics performance compared with previous-generation processors, according to Advantech. This is intended to support graphics-intensive workloads such as 3D applications, AI-assisted training and high-resolution content rendering without requiring an external GPU.
The motherboards are designed to support workloads including real-time defect inspection, predictive maintenance and intelligent monitoring directly at the edge. Keeping processing closer to equipment can help reduce dependence on cloud connectivity while supporting quicker responses from industrial systems.
Advantech also highlights software support through its EdgeAI SDK, including GenAI Studio for custom large language model applications. Its orchestration platform can further support fleet-wide updates, helping organisations manage deployed edge systems.
By combining Intel 18A-based processing with integrated CPU, GPU and NPU capabilities, the new platforms strengthen the hardware foundation for industrial AI. The focus is on enabling manufacturers to process data locally while supporting increasingly demanding vision, automation and intelligent-monitoring applications.
