Teaming Up with Mobilint to Accelerate Edge AI Adoption

POSCO DX is accelerating the transformation of its artificial intelligence (AI) infrastructure from foreign-made GPU-centric systems to domestically developed NPU-based systems, thereby revolutionizing the application of AI on manufacturing sites. The core aim is to implement 'Edge AI,' which not only reduces infrastructure costs but also ensures both security and real-time processing.

POSCO DX AI Research Institute is equipping its control system PosMaster with an NPU to test AI models in industrial sites. POSCO DX

POSCO DX AI Research Institute is equipping its control system PosMaster with an NPU to test AI models in industrial sites. POSCO DX

View original image

According to POSCO DX, the company signed a business agreement on April 2 at its Pangyo office with AI semiconductor startup Mobilint to implement AX (AI transformation) using NPUs. Previously, in February, POSCO DX, together with POSCO Technology Investment, invested a total of 3 billion won in Mobilint through the POSCO DX Corporate Venture Capital (CVC) New Technology Investment Association.


An NPU is a semiconductor specialized for AI computation such as deep learning and machine learning. Compared to GPUs, NPUs are optimized for AI inference, which can lower infrastructure costs and improve power efficiency. In particular, by integrating NPUs directly into equipment control systems, 'Edge AI' can be realized without routing data through remote data centers, enabling real-time analysis and control on manufacturing floors without transferring data externally.


Through this collaboration, the two companies will apply Mobilint’s NPU to POSCO DX’s industrial control system, PosMaster. Based on this, they plan to create an intelligent factory that enables immediate AI analysis and control at the equipment control stage. Additionally, they will jointly develop technology to ensure POSCO DX’s AI model algorithms deliver optimal performance in Mobilint’s NPU environment.


Mobilint is a domestic company with high-performance NPU technology capable of running large language models (LLMs) in edge environments, allowing for real-time, on-site data-driven decision-making and control. This is expected to further accelerate POSCO Group’s construction of intelligent factories.


Going forward, POSCO DX plans to expand its Edge AI technology, shift its AI systems from GPU-based to NPU-centric architectures, and broaden its application to various industrial sites, including steel and secondary battery material production as well as logistics.



Jo Seokjoo, Head of the AX Convergence Research Institute at POSCO DX, stated, "Through this partnership, we have secured LLM-based Edge AI technology, which is essential for realizing intelligent factories," adding, "By collaborating with domestic startups, we will lead the paradigm shift in manufacturing AI."


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing