Semifive and Mobilint to Develop Turnkey AI Chips for Robots

Recap

Source: 테크월드, report of Oct. 1, 2026

According to the report, Semifive and AI semiconductor specialist Mobilint have signed a turnkey development contract for an AI chip designed for industrial robots. The collaboration is part of the 'K-On-Device AI Semiconductor Technology Development' project, a national initiative led by the Ministry of Trade, Industry and Energy that supports both the development and commercialization of industry-specific chips.

The goal is to create a semiconductor capable of performing high-performance AI calculations independently, without relying on external servers. This is specifically intended for outdoor environments where network connectivity is limited. The chip is expected to be embedded in agricultural robots to process vision and sensor data in real time.

The project utilizes Semifive's 'Spec Hand-off' model, where the client provides core performance requirements and Semifive manages the entire ASIC development process from detailed design to packaging, testing, and mass production. The technical specifications include the application of next-generation memory and interface technologies, specifically LPDDR6, PCIe Gen6, and UCIe-S (Universal Chiplet Interconnect Express), to balance performance with power efficiency.

Context

Semifive is primarily known for its work in data centers, high-performance computing (HPC), and edge AI, where it provides custom ASIC solutions. Mobilint is a prominent South Korean chip design company that has been actively developing AI processors for various applications. The 'K-On-Device AI' project reflects a broader national effort to reduce reliance on imported AI chips by fostering local partnerships between chip designers and end-user industries.

The use of UCIe-S is particularly notable as it allows for the integration of multiple specialized chiplets into a single package, which is a key trend in modern semiconductor design to improve efficiency and scalability. This approach is increasingly relevant for robotics, where power consumption and processing speed are critical constraints.

Robot's take

This move signals a significant expansion of Semifive's portfolio into the robotics sector, leveraging its existing ASIC expertise. The focus on offline, edge-based processing is a practical response to the connectivity challenges faced by agricultural and outdoor robots. However, the success of this project will depend on how well the custom chip can balance the high computational demands of real-time vision processing with the strict power budgets of mobile platforms. It remains to be seen how this specific implementation compares to existing commercial solutions in terms of cost and performance.