Neurocle, a specialized AI vision software company, unveiled its auto deep learning vision inspection technology optimized for the latest Intel® Core™ Ultra Series 3 and Intel® Core™ Series 2 platforms at the Intel Edge Solution Summit 2026.

Through a technical seminar, panel discussion, and live demonstrations at the event, Neurocle presented a new direction for Edge AI, demonstrating how manufacturing sites can build AI vision inspection systems more easily and efficiently.

■ Optimized for Latest Intel Platforms: High-Performance GPU-less AI Vision Inspection

During the event, Neurocle CEO Hongsuk Lee hosted a technical seminar to share the optimization process of its auto deep learning vision inspection software on the Intel® Core™ Ultra Series 3 and Intel® Core™ Series 2 platforms.

The presentation highlighted how the software efficiently utilizes CPU, iGPU, and NPU resources through Intel OpenVINO™-based optimization, alongside actual vision inspection performance results achieved on the latest Intel processors.

Notably, on the Intel® Core™ Ultra Series 3 platform, the software achieved up to a 4.76x increase in inference performance for Classification models and up to a 4.39x increase for Segmentation models compared to the previous generation Intel® UHD Graphics 770, while the NPU also demonstrated outstanding AI processing performance matching the capabilities of an iGPU.

With these results, Neurocle proved that the level of AI vision inspection required on the factory floor can be fully implemented without the need for separate, high-cost GPUs.

■ "Model Accuracy Alone is Not Enough": Presenting Optimization Strategies for Manufacturing AI Expansion

  CEO Hongsuk Lee also participated in a panel discussion on AI Transformation alongside global companies including Intel, IGEL, Iternal Technology, and Trossen Robotics.

Representing the manufacturing vertical, Lee shared insights on how manufacturers can successfully adopt and scale AI in production environments, emphasizing that deployment simplicity, hardware optimization, and operational scalability are becoming just as important as model accuracy.

Neurocle provides optimization across Intel platforms based on 'Neuro-R,' its proprietary AI runtime library integrated with Intel OpenVINO™. This allows manufacturers to stably run identical AI models without separate model redevelopment or additional optimization, even during hardware changes or platform upgrades.

In particular, Neurocle’s standardized AI software optimized for Intel platforms reduces the integration burden of AI solutions for Industrial PC (IPC) manufacturers, while providing a simpler and faster AI deployment environment for manufacturing end-users.

■ Core Ultra Series 3·Core Series 2: Real-Time Vision Inspection (Automotive·Pharmaceutical)

At the event, Neurocle showcased real-time vision inspection demonstrations for the automotive and pharmaceutical industries running on the latest Intel platforms.

The demonstration on the Intel® Core™ Series 2 platform showcased high-resolution image processing performance based on up to 12 Performance Cores through an automotive wheel inspection demo, while the demo on the Intel® Core™ Ultra Series 3 platform showcased advanced iGPU and NPU-based Edge AI inference performance through a pharmaceutical quality inspection setup.

These live demonstrations were evaluated as a practical case study proving that actual AI vision inspection systems can be built on the factory floor without relying on expensive GPU infrastructures.

Through this event, Neurocle emphasized that the competitiveness of AI vision inspection is shifting beyond simple model accuracy toward ease of deployment, scalability, and operational efficiency.

"What manufacturing sites truly need is not just a high-accuracy AI model, but an AI solution that can operate stably on an actual production line," said Neurocle CEO Hongsuk Lee. "Through our collaboration with Intel, we are building a standardized Edge AI environment that enables both device manufacturers and IPC partners to adopt AI more easily and rapidly."

Meanwhile, Neurocle plans to expand its collaboration with Intel and global IPC partners to continuously advance CPU, iGPU, and NPU-based Edge AI optimization technologies, while continuing technical development to lower the barrier to entry for AI adoption on the factory floor.