- Integrated call for key projects for supporting AI semiconductor demonstration from May 29 (Thu) to June 30 (Mon)

​

The Ministry of Science and ICT (“MSIT”; Minister Yoo Sang-im) announced that it will launch a call for major new projects in the field of AI semiconductor, implemented with a newly allocated supplementary budget to support the early commercialization of neural network processing devices currently being launched by domestic fabless companies, from Thursday, May 29 to Monday, June 30.

​

The MSIT has allocated a supplementary budget of KRW 49.4 billion this year to provide practical support to promising AI semiconductor companies. The main goals are as follows: 1) Enhancement of AI computing demonstration infrastructure (KRW 12 billion); 2) Development and demonstration of “Device AX”—AI transformation of domestic AI semiconductor-based devices (KRW 6 billion); 3) Support for AI transformation (AX) demonstration (KRW 4 billion); 4) Support for overseas demonstration of AI semiconductors (KRW 5.4 billion); and 5) Timely support for AI semiconductor commercialization (KRW 22 billion). The MSIT will invest a total of KRW 243.4 billion—including the supplementary budget—throughout the entire cycle of AI semiconductor development, from R&D to demonstration to workforce development, to boost the AI semiconductor industry.

​

The projects launched with the supplementary budget were carefully planned, reflecting the private sector’s demand gathered through various meetings and discussions prior to securing the budget. They were designed to directly support fabless companies‘ activities, such as early commercialization of neural network processing units (NPUs). The Ministry is launching a call to recruit companies and institutions to carry out the following projects: (1) AI computing demonstration infrastructure upgrade project, (2) AI transformation demonstration support project, and (3) AI semiconductor commercialization timely support project.

​

< Call for New Projects >

​

​

First, the AI computing demonstration infrastructure enhancement project will be carried out over three years, shifting from the existing small-scale, technology verification-focused approach to large-scale domestic NPU demonstration to support early commercialization. To this end, a dedicated cluster with a capacity of 120 petaflops will be established, and a demonstration infrastructure compatible with commercial AI computing services will be created. The project will start with building a 50-petaflop infrastructure capable of demonstrating commercial large language models (LLMs) this year; in 2026, it will be gradually expanded and full-scale pilot demonstration of diverse LLMs will begin.

​

​

Second, the AI Transformation Demonstration project will be carried out for two years, supporting the conversion of various commercial AI services available on the market into domestic AI semiconductors, as well as their commercialization. The project will support the conversion of four commercial AI services linked to local industries, promoting the AI transformation of the industry and spreading it nationwide to promote the NPU-powered AI inference market.

​

​

The last project for providing timely support for AI semiconductor commercialization is aimed at addressing the current shortage of support measures; its goal is to establish a comprehensive support system for promising AI semiconductor startups throughout the entire cycle of commercialization—from design software to product manufacturing to card/server-level verification. In particular, the project will maximize practical support through vouchers and other methods to enable companies to receive support tailored to their specific circumstances. Moreover, customized consulting and technical support will be provided at each stage to help promising fabless companies enter the market swiftly.

​

This call will recruit companies eligible for the support for prototype and mass production. The recruitment for companies in need of design software vouchers will start in June. Furthermore, by the end of this year, enhancements will be made to the AI semiconductor verification system in the Gwangju AI Cluster to provide card/server-level stability, compatibility, and reliability verification essential for fabless companies' product commercialization.

​

Moreover, the AI semiconductor overseas demonstration support project, which received an additional budget of KRW 5.4 billion, was announced on May 16 in advance, considering the preparation time necessary for its implementation (e.g. consultation with overseas demand sources). A separate new call will be launched in June to recruit companies for the development and demonstration project for AI transformation of devices equipped with domestically manufactured AI chips.

​

Director General Park Tae-wan of the ICT Industry Policy Bureau at the MSIT said, “The MSIT has recognized the importance of AI semiconductors and has supported the domestic AI semiconductor industry by identifying policy needs through close communication with companies and commercializing their products. We will continue to work as one team with the industry to ensure that domestically manufactured AI semiconductors are commercialized in a timely manner and that companies receive practical support.”

​

For further information, please contact the Public Relations Division (Phone: +82-44-202-4034, E-mail: msitmedia@korea.kr) of the Ministry of Science and ICT.

​

Please refer to the attached PDF.