알림 목록

Analysis Area: Daegu Metropolitan City

Core Area: Suseong Alpha City, DGIST, Daegu Technopark, High-Reliability Semiconductor Development Support Hub, and Mobility, Medical, and Robotics Enterprise Zone

Agenda:  Is Daegu's semiconductor strategy moving beyond scale competition with the Seoul metropolitan area to secure specialized achievements in high-reliability semiconductors for mobility, medical, and robotics?

Golden Time Type: Specialization Window + Scale Risk

Reference Date: August 28, 2026

Version: Regional AX Golden Time Intelligence v3.2
 

Daegu designated semiconductors as one of its five new industries and entered into support for high-reliability semiconductors, AI semiconductors, sensors, and fabless companies, despite lacking large-scale memory fabs or established semiconductor supply chains. With the Seoul metropolitan area possessing design firms and large demand companies, the Chungcheong region having production facilities, and Gyeongbuk having a foundation in materials, components, and the electronics industry, Daegu has limited room to enter into a scale competition in the general-purpose semiconductor sector.  Recent industrial statistics compiling the sales, employment, and mass-produced products of semiconductor companies in Daegu are limited, and if specialized markets are not secured through actual purchasing contracts over the next two to three years, only support facilities will remain.  The viability of Daegu's semiconductor strategy is determined not by the scale of the industry, but by whether it can preemptively secure high-reliability demand markets.

Daegu possesses automotive parts, robots, medical devices, and AI as strategic industries, creating a structure capable of demonstrating local demand for automotive, robotic, and medical semiconductors . However, the existence of demand industries is a separate issue from the purchase of local semiconductors, and  evidence linking the demonstration, certification, mass production, and long-term supply contracts of chips developed by local companies is still limited. Since  the reliability verification period for automotive and medical semiconductors is long, if design and demand companies are not aligned between 2026 and 2028, revenue generation will be delayed beyond 2029.  Although Daegu has established a direction for specialization, the timeline for market entry has not yet been secured.

Daegu is implementing a support project for the verification and validation of high-reliability semiconductor commercialization, investing a total of 21.225 billion won from 2025 to 2029. With 14.475 billion won in state funding, 6 billion won in municipal funding, and 750 million won in private investment, the project establishes a development support center, verification equipment, design tools, verification IP, and a prototype evaluation infrastructure. This move solidifies the regional strategy from fostering general semiconductors to supporting fabless companies focused on safety and reliability.  However, private investment accounts for only about 3.5% of the total project cost, and the scale of mass production orders and self-investment by participating fabless companies remains unconfirmed. If public verification demand does not shift to private mass production investment by 2028, the center will remain merely a research support facility. While specialization in high reliability has begun, it is premature to classify it as an industry with a confirmed market.

The D-FAB Sensor Foundry will invest 31 billion won in state funds from 2023 to 2027 to construct a 6,000-square-meter facility comprising cleanrooms, sensor production facilities, and corporate support spaces . While Daegu is shifting from a region providing only design support to one capable of small-batch sensor manufacturing and prototype validation, D-FAB is a shared infrastructure for development and verification, not a mass production fab.  The contractual structure with external foundries and packaging/testing companies responsible for mass production following the prototype stage is not confirmed in the publicly available data. If a mass production pathway is not established after 2027, companies will have to seek production networks in the Seoul metropolitan area, the Chungcheong region, or overseas again following the success of their prototypes. Although  Daegu is securing a foundation for sensor semiconductor prototypes, it does not yet constitute a complete production ecosystem.

Companies such as the system semiconductor design and IP firm Ideas2Silicon, the infrared sensor company ST Labs, and the research organization of the automotive semiconductor company Telechips have entered Daegu. The scale of investment agreements signed by five semiconductor-related companies in Daegu from 2023 to 2025 was tallied at 526.7 billion won, and the design and workforce bases of Kyungpook National University and DGIST have also expanded. However, the investment figures include both the semiconductor chips themselves and related industries such as printed circuit boards, and separate data regarding actual execution amounts, dedicated semiconductor employment, and regional revenue are unavailable.  If the headquarters, research institutes, and mass production functions of the signatory companies do not settle in as planned over the next two to three years, the gap between the scale of investment attraction and the industrial ecosystem will widen.  While the influx of companies has begun, the density of integration remains low.

The Intelligent Semiconductor Development Support Center provides EDA tools, MPW, test analysis, and performance verification, while the High-Reliability Semiconductor Development Support Center adds functional safety and reliability verification systems. Although a foundation has been established to allow design firms to reduce initial development and verification costs,  Daegu lacks commercial fabs responsible for mass production of advanced processes.  If production and back-end processes relocate to external regions after design and verification, the pathway for technology development achievements to translate into local production and employment is severed. Unless regular MPW and mass production slots with external foundries are secured by 2028, local fabless companies will once again be exposed to risks regarding costs and delivery times after public support ends.  Daegu is developing into a design and verification hub rather than a semiconductor production base.

D-FAB is responsible for the small-batch production and development of sensor semiconductors, while DGIST operates a 6-inch CMOS design verification platform and a 'My Chip' fabrication service. Although accessibility to fabrication at the research and prototype levels has improved, advanced AI semiconductors and high-performance automotive chips require different process, packaging, and reliability conditions.  There is no evidence that regional infrastructure encompasses all specialized semiconductors. If the scope of support expands over the next two to three years, there may be a mismatch between equipment utilization and actual corporate demand.  Daegu's fabrication base is suitable for sensors and verification, but it is not a fab base that covers the entire spectrum of AI and automotive semiconductors.

Semiconductors for automobiles, medical devices, and robots require higher reliability regarding temperature, vibration, error, lifespan, and functional safety than chips for general consumers. Daegu possesses automotive, medical, and robotics industries, enabling it to provide a testing environment for high-reliability semiconductors; however,  instances where local demand companies have proposed specifications and committed to purchase from the early stages of development are limited.  Even if prototype verification is successful, revenue will not be generated unless the product is adopted in the designs of automakers, medical device manufacturers, or robot companies. If collaboration with demand companies remains limited to demonstration agreements until 2028, specialization in high reliability will remain merely a technical classification and fail to establish itself as a market classification.  While the rationale for Daegu's specialization holds, the buyer base remains undetermined.

By 2026, Daegu is expanding cooperation with universities, fabless companies, and support organizations for the demonstration and commercialization of domestic AI semiconductors, and connecting the robotics and mobility sectors to demand industries. Although a structure has been established to combine AI semiconductors with the local AX infrastructure , the performance of the chips being demonstrated, costs compared to imported products, and the scale of purchasing conversion and subsequent mass production have not been disclosed. If domestic chips are used only for demonstration purposes for two to three years and are not adopted in the basic designs of commercial products, the reference effect will be limited. While AI semiconductor demonstration has begun, market verification has not yet been completed.

Telechips pursued the establishment of a research institute in Daegu and the expansion of its specialized workforce, while Ideas2Silicon and ST Labs also established bases in the city. The entry of external fabless companies into the region serves as evidence that a corporate foundation is being created in Daegu, a city that was once a semiconductor wasteland; however,  the extent to which headquarters functions, core design rights, IP ownership, and revenue attribution remain in Daegu varies by company.  If only small-scale research organizations are deployed locally, business decisions and high-value-added profits will remain in the headquarters' region, even if the number of companies increases. If regional bases fail to secure independent products and revenue responsibilities by 2028, Daegu will remain in a position to provide dispersed research personnel. While  the attraction of companies has progressed, their internalization into the local industry remains unconfirmed.

The 526.7 billion won in investment agreements from five companies between 2023 and 2025 includes Isu Petasis's investment in high-layer printed circuit boards. While it is clearly a large-scale investment linked to the AI ​​semiconductor supply chain, PCB production is an industrial stage distinct from semiconductor design and wafer manufacturing.  Integrating investments in related industries into the semiconductor industry's performance results in an overestimation of the region's actual fabless, design, and sensor bases. Unless statistical classifications become more sophisticated over the next two to three years, it will be difficult to distinguish between increased investment and the growth of core semiconductor competencies. Although  the scale of semiconductor investment in Daegu has increased, the composition of the value chain remains unclear.

Kyungpook National University has established a foundation to produce approximately 140 talents annually by adding 100 admission slots in the field of next-generation intelligent semiconductors and including 40 students in convergence majors . DGIST has also rapidly expanded its regional educational supply by operating a Department of Semiconductor Engineering and a design verification platform. On the other hand , the total hiring scale of fabless and design firms in Daegu, as well as the local employment and two-year retention rates of graduates, remain unconfirmed. If the corporate base grows more slowly than the educational supply by 2028, trained personnel will migrate to production hubs in the Seoul metropolitan area and the Chungcheong region. While  Daegu possesses the infrastructure to cultivate semiconductor talent, the industrial base to absorb it within the region is still small.

Kyungpook National University operates an AI semiconductor convergence major and industry-linked PBL, and job postings from local companies are also being integrated with university education. While the curriculum covers materials, processes, devices, and design, corporate demand in Daegu is concentrated in the fields of design, verification, and sensors; therefore, the scope of education may not perfectly align with the local job structure. If employment outcomes over the next two to three years are tallied solely based on total semiconductor employment, the contribution to the local industry cannot be determined.  Although the quantitative foundation for workforce training is at an above-average level, its alignment with local job roles remains unconfirmed.

Daegu has designated future mobility, robotics, and healthcare as key industries and positioned semiconductors as the common foundation for them. While automotive electronics and sensors, robot control and edge AI, and medical imaging and diagnostic devices generate demand for high-reliability semiconductors,  sales figures for semiconductors purchased by local demand companies from local design firms have not been disclosed.  Unless demonstration projects and technology agreements translate into purchase contracts, the benefits of demand industry clustering remain merely declarations. If the first mass-produced products do not emerge by 2028, market credibility for the region-linked specialization strategy will decline.  While technological connections between industries have been established, commercial connections are still in their early stages.

The Regional Hub AX Innovation Technology Development Project includes a research and development structure that combines robotics and biotechnology with AI and semiconductors. While public R&D bears the costs of inter-industry linkage,  a structure in which private companies repeatedly purchase the same semiconductors after the project ends has not been confirmed. If demand is formed solely by project funding for two to three years, semiconductor companies cannot verify actual market prices and volumes.  Daegu's demand linkage is based on public demonstration, and there is insufficient evidence to classify it as a self-sustaining supply chain.

The High-Reliability Semiconductor Verification Support Center has secured the status of being the first of its kind outside the capital region, and the D-FAB also possesses the rarity of a regional sensor foundry. However, since the center is open to all non-capital fabless companies, external firms can also utilize Daegu's infrastructure, and  there is no confirmed obligation to maintain production, headquarters, or revenue in Daegu following verification. If utilization rates by external firms remain high while the establishment and investment of local companies are low by 2029, facility performance and regional industrial performance will be separated.  While Daegu possesses scarce public infrastructure, the structure does not ensure that its utilization value accrues exclusively to the region.

Fabless companies, venture capital firms, and large-scale demand companies are concentrated in the Seoul metropolitan area, while production bases for memory, foundry, and packaging have been established in the Chungcheong region. As Daegu builds a miniature version of the same industry, existing clusters are also expanding their scope into AI, automotive, and sensor semiconductors. Unless the number of Daegu-specialized companies, revenue, and mass production cases increase rapidly, the time advantage of being the "first outside the capital" will weaken within two to three years.  Currently, Daegu's differentiation lies in securing facilities, but it has not yet been confirmed as securing the industry.

High-reliability semiconductors (21.225 billion KRW), D-FABs (31 billion KRW), and the Intelligent Semiconductor Development Support Center and university design verification platforms have been identified. While infrastructure for design, prototyping, and verification is under development, there is a shortage of large-scale fabs, back-end process integration, mass production orders, and a large number of fabless companies. If the number of design firms and demand companies does not increase by 2028, the supply capacity of support facilities will outpace the demand of local industries. The level of readiness is assessed as upper-middle for public infrastructure, lower-middle for the corporate ecosystem, and lower for the mass production system .

The entry of fabless and sensor companies, such as the Telechips research base, Ideas2Silicon, and ST Labs, has been confirmed. However, data comparing the total number of semiconductor companies in Daegu, proprietary products, sales, and follow-up investments in a continuous manner is limited.  Industrial clustering will not be formed unless startups, spin-offs, and supplier companies expand beyond the small number of companies attracted over the next two to three years. The level of company expansion is assessed as "initial attraction in progress" and "ecosystem expansion unconfirmed ."

The direction of AI semiconductor agreements, robot and mobility demonstrations, and the application of high-reliability semiconductors in medical and automotive sectors is confirmed. On the other hand, the purchase amounts, product adoption, and mass production contracts of local demand companies were not disclosed in an integrated manner. If demonstration chips are not embedded in commercial products by 2028, specialized demand industries will remain merely as testbeds for semiconductors. The level of demand industry diffusion is assessed as entry into demonstrations and adoption in mass production remaining unconfirmed .

Kyungpook National University’s foundation for producing approximately 140 semiconductor talents annually and DGIST’s foundation for design and process research are confirmed. On the other hand, the private sector contribution to the high-reliability project amounts to only 750 million won, and venture investment and scale-up funding for local fabless companies are confirmed to be limited. If corporate investment and hiring fail to keep pace with the supply of education for two to three years, the workforce will migrate out of the region, and the utilization of public equipment will rely on external companies. The readiness levels for manpower and capital are assessed as upper-middle for education and lower-middle for private capital .

The D-FAB project will run until 2027, and the high-reliability semiconductor business until 2029; the regional hub AX project will also promote the integration of robot, bio, and AI semiconductors during the same period. The period from 2026 to 2028 is intended to complete the design, prototyping, verification, and adoption by client companies of the first product, but currently, Evidence is focused on facilities and agreements. If mass production orders do not materialize during this period, companies will remain in the early development stage even after the center is completed in 2029. The golden time for Daegu's semiconductor specialization strategy is assessed to be the next 24 to 30 months .

D-FAB and the two development support centers maintain cleanrooms, equipment, EDA tools, and operational personnel for the long term.  Even if the supported fields align with actual corporate demand or mass production pathways are not secured, costs for equipment maintenance and replacement, as well as specialized personnel, continue to accrue. If local company utilization and commercialization rates remain low by 2028–2029, support for external companies to operate the facilities will expand, and regional specialization will weaken. Irreversibility is assessed as high risk for facility investment, medium risk for workforce outflow, and high risk for market entry .

Daegu's semiconductor strategy has materialized by avoiding competition in general-purpose production and connecting high-reliability, sensor, and AI semiconductors with local demand industries.  The logic behind this as a specialized strategy for non-metropolitan areas holds true.

However, the current focus of the industry lies in public development and verification infrastructure rather than companies and products .  There is a significant gap between mass production fabs, back-end processes, private capital, and recurring purchase agreements.

The final assessment is “High-Reliability Niche Positioning + Commercial Scale Gap .  ” While the Daegu semiconductor strategy has secured a viable specialized direction, it is not yet at the stage where it can be judged to have survived as an industry.

Overall Score: 53 / 100

Golden Time Status:
ORANGE–RED

Time Window:
24~30 months

Failure Mode:
The first development and verification facility outside the capital area remains, but design firms, mass production revenue, and highly skilled personnel relocate to the capital area and existing production bases.

Irreversibility:
High risk of facility investment and market entry

Evidence Sources

• Daegu Investment Environment—Establishment of D-FAB Sensor Foundry
• Intelligent Semiconductor Development Support Center
• Daegu Industrial and Economic Trends—Daegu's 5 Major New Industries
• Daegu High-Reliability Semiconductor Commercialization Verification and Confirmation Support Project
• DGIST—Semiconductor Design Verification and 6-inch CMOS Platform
• DGIST Department of Semiconductor Engineering
• Kyungpook National University—Multilateral Agreement for Fostering the AI ​​Semiconductor Industry
• Kyungpook National University—Next-Generation Intelligent Semiconductor Enrollment and Convergence Major
• DGIST—Advanced Sensor and System Semiconductor Platform for AI, Robotics, and Mobility

Structural Insight — Daegu Semiconductor's Actual Product is Closer to 'Reliability Proof' Than 'Chip'

Daegu lacks advanced mass production fabs, and it is difficult to compete with existing semiconductor clusters in terms of production scale in the short term. On the other hand, demand industries with high error costs, such as automobiles, robots, and medical devices, as well as high-reliability verification centers and sensor manufacturing bases, are being established in the same region.  The scarce asset secured by Daegu is not wafer production volume, but rather a proven history of chips operating safely in specific environments.

For high-reliability semiconductors, market value is established not at the moment of design completion, but after a long period of verification and product adoption. Even for identical chips, once performance and safety are confirmed in actual products such as automobiles, medical devices, or robots, trust assets are accumulated that are difficult for latecomers to replicate in the short term.  These assets create entry barriers that differ from the scale of production facilities.

Therefore, Daegu’s specialization strategy is not about competing by miniaturizing semiconductor production cities, but rather about ensuring survival within a structure where local demand industries prove the reliability of the chips. However, if the results of the demonstration do not lead to purchase contracts and mass production history, the verification ends with public test reports. The ultimate competitiveness of Daegu’s semiconductor industry is determined not by how many chips were designed, but by how many actual products in Daegu have proven the reliability of those chips .