South Korea’s Southwest Chip Hub: The 2030 Plan and Its Biggest Tests

Southwest South Korea’s proposed semiconductor hub has a production target: 2030. At the September 2026 Jeonnam–Gwangju Semiconductor and AI Transformation Global Forum, participants argued that meeting it would require far more than building fabs. Power, water, suppliers, research capacity and people would need to come together on a similar timetable.13

The strategy: build an ecosystem, not an isolated fab

The proposal aims to establish a second major chip production base beyond the Seoul metropolitan area by combining Gwangju’s AI and advanced-manufacturing strengths with South Jeolla’s land and energy resources. Forum discussions envisioned links among fabrication, advanced processes, AI, materials, components, equipment, packaging, research and workforce development.13

That breadth matters to the project’s purpose. A factory could bring production to the region; a network of local suppliers, researchers and trained workers would give it a stronger foundation for continued operations and expansion.3

The timetable—and what it does not guarantee

Mayor Min Hyung-bae set out a first-phase schedule: begin construction in 2027, finish by the end of 2029 and start initial production in 2030. These are targets, not confirmation that the facilities and their supporting infrastructure are secured.3

Jung Won-ju, chair of Herald Media Group, Jungheung Group and Daewoo Engineering & Construction, pledged 10 billion won to support development of the semiconductor and AI cluster. It is a contribution to the effort, not a financing plan for the entire hub.2

The decisive test: infrastructure that reaches the fabs

The forum put timely electricity, industrial water, permits, transport, logistics and supplier networks at the center of execution. Planning cannot stop at the first buildings: land and infrastructure must also leave room for expansion beyond the initial four-fab concept.3

Power illustrates the distinction between regional potential and a working production site. One estimate put the electricity-receiving capacity for four fabs at 6.4 gigawatts. Even with a regional generation surplus, grid reinforcement and stable delivery to the sites would still be necessary. The estimate describes a planning requirement, not power already available at the fabs.3

Workers, chipmakers and the Kumamoto lesson

Participants also emphasized developing workers, suppliers and research institutions within the region. Training and R&D would need support from housing, healthcare, education and cultural amenities that help skilled workers settle locally.3

TSMC’s cluster in Kumamoto, Japan, offered a lesson in government support and local coordination, but not a template to copy wholesale. The southwest Korean project must fit Samsung Electronics’ and SK hynix’s actual investment, production and supplier requirements.3 The 2030 goal therefore rests on coordination: factories, utilities and a durable local industrial community must become ready together.3