Intel's Edge AI Controller and All-in-One IPC

At the 2025 VisionChina exhibition in Beijing, Intel and NODKA unveiled an edge AI controller and an all-in-one edge computing system, leveraging Intel? Core? Ultra 200H processors and Intel? Arc? Pro GPUs. These solutions enhance industrial AI applications, driving smart manufacturing with high performance and efficiency.

Edge AI Controller: NP-6125-H1B

The edge AI controller (NP-6125-H1B) addresses the demand for cost-effective, compact AI solutions in industrial settings, such as machine vision, robotic guidance, and embodied intelligence. Powered by Intel Core Ultra 200H processors, it overcomes the high power consumption and cost of traditional 4U industrial PCs. Key features include:

• Integrated SoC Architecture: A single SoC design eliminates reliance on discrete GPUs, reducing power consumption and enabling fanless operation for cost-effective, reliable deployment.
• High Edge AI Performance: Integrates a neural processing unit (NPU) and integrated GPU (iGPU) for complex model inference, ensuring real-time image processing and task analysis with high accuracy.
[](https://newsroom.intel.com/client-computing/2025-ces-client-computing-news)
• Scalability and Flexibility: Supports DDR5 memory (up to 6400 MT/s) and modern interfaces, offering upgrade paths for evolving technological needs.
• Software Ecosystem: Intel? CVOI provides optimization guidelines and code samples, while Intel? OpenVINO? supports seamless integration across Intel hardware and deep learning frameworks, enhancing vision system performance.

The controller¡¯s compact design and low power consumption make it ideal for flexible industrial deployment.

Edge Computing All-in-One: IPC-615H5 and AIPC-3000-AQ6-7L

The edge computing all-in-one system, available in 4U (IPC-615H5) and compact single-GPU (AIPC-3000-AQ6-7L) configurations, targets complex AI workloads. Powered by 12th Gen or later Intel Core processors and dual Intel Arc Pro B60, A770, or B580 GPUs, it delivers up to 197 TOPS (INT8) with 24GB VRAM per GPU. Applications include intelligent customer service, equipment maintenance, and industrial code generation.

• High AI Performance: Handles large-scale language models, multimodal models, and high-precision image/video analysis.
• Fault Prediction: Enables proactive identification of system issues, reducing downtime and maintenance costs.
• Compact Design: The single-GPU model suits space-constrained environments without sacrificing performance.

Market Context and Impact

China¡¯s industrial AI spending is projected to reach 900 billion RMB by 2028 (CAGR 37.7%), with vision-based industrial PCs comprising over 60% of the market in 2024, expected to grow further. These solutions address the need for high-performance, low-power edge AI, enhancing production efficiency, safety, and smart manufacturing.

Conclusion

The Intel-NODKA edge AI controller (NP-6125-H1B) and all-in-one systems (IPC-615H5, AIPC-3000-AQ6-7L) leverage Intel Core Ultra 200H processors and Arc Pro GPUs to deliver scalable, efficient AI solutions for industrial applications. Supported by a robust software ecosystem, these systems enable real-time vision processing and complex AI tasks, driving intelligent transformation across industries like manufacturing, retail, and aerospace.