Taipei, March 9, 2026 /PRNewswire/ — DFI, a global provider of industrial computers, embedded motherboards, and edge computing platforms, announced that at Embedded World 2026 it will showcase an application-driven edge AI platform developed in collaboration with Intel, highlighting how edge AI is moving from pilot demonstrations to scalable real-world deployments across industrial and infrastructure environments.
As edge AI moves beyond proof-of-concept use, industrial customers are becoming more focused on platforms that balance performance, power efficiency, and long-term uptime. DFI said it is addressing these requirements through a layered edge AI product portfolio designed to support deployment in harsh industrial conditions, with a robotic arm application developed in collaboration with Intel at its core.
At Embedded World 2026, DFI will unveil a new generation of mission-critical edge AI platforms for defense and medical applications. For defense and unmanned systems, DFI introduces the PTH9HM COM-HPC Mini module. It is a credit card-sized, SWaP-optimized compute engine that enables real-time 8K vision processing and edge AI inference for autonomous navigation, target recognition, and threat tracking. Powered by an Intel Core Ultra Series 3 processor with integrated Arc GPU, the PTH9HM supports wide temperature operation from -40 °C to 85 °C, integrates 64 GB of onboard LPDDR5x memory, and features PCIe Gen 5, dual 2.5 GbE and TPM 2.0 security to protect mission-critical data.
For medical imaging and diagnostics, DFI also introduces the PTH171/173 Edge AI Mini-ITX motherboard. Featuring an Intel Core Ultra Series 3 processor that delivers up to 180 total TOPS with Intel Arc graphics, PCIe Gen 5 expansion, rich multi-display and I/O connectivity, and Intel vPro management, the motherboard is designed for long-lifecycle medical deployments that require high precision and reliability.
The SF101-PTH industrial performance compact system is powered by an Intel Core Ultra Series 3 processor. The system’s heterogeneous architecture integrates Intel CPUs with integrated GPUs and NPUs to support real-time control, vision processing, and AI inference within a single edge platform.
The solution uses integrated AI acceleration and real-time technologies such as Intel Time Coordination Computing (TCC) and Time-Sensitive Networking (TSN) to achieve millisecond-level responsiveness without a separate GPU, optimizing performance per watt while reducing total cost of ownership (TCO), and forming a scalable foundation for physical AI and robotics applications.
