KR — 전자신문
National AI Strategy Committee Visits Jeonnam-Gwangju AI and AX Sites, Pledges Megaproject Support
Policy1 min read
South Korea's National AI Strategy Committee visited the Jeonnam-Gwangju region for two days to look at AI chip, AI infrastructure, and shipbuilding AI sites. The trip supports the government's plan to turn the region into a key chip and AI hub.
What happened
On the 19th, the committee began a two-day visit to Jeonnam-Gwangju. Its deputy chair, Ha Jung-woo, met with local companies that make Korean AI chips. On the 20th, the team toured HD Hyundai Samho's shipyard in Yeongam and a national AI computing center under construction in Haenam.
Key numbers and facts
- Seven Korean neural processing unit (NPU) companies joined the chip meeting.
- The national AI computing center will cost about 2.5 trillion won (about 2.5 trillion won) and have 40 MW of power capacity.
- The center is planned to open for full service in 2028.
- The government plans to add four memory fabs in Jeonnam-Gwangju, making it a second chip production hub.
Why this region matters
In June, the government named Jeonnam-Gwangju as a place for new memory chip factories as part of its three big mega projects. In August, it said the southwest region would grow into a clean-energy-based chip and AI cluster. This visit is a follow-up to those plans.
Why it matters
Korean AI chips need real tests and first customers to grow. The companies said it is hard to prove their chips work in real services and to win early users. The government says it will help with testing, public and private use, and sales. The shipyard visit also shows how AI and robots could fix labor shortages in heavy industry.
What to watch
- Progress on the NPU testing center run by AICA and TTA.
- Plans for a computing center built just for NPUs, led by the Ministry of Science and ICT.
- Whether the K-Perf chip performance test is used to check Korean AI chips.
- Power and water supply for the Haenam data center, and whether solar power is used.
Glossary
- NPU
- A chip built to run AI tasks, like a brain for computers that learn from data.
- AICA
- The Artificial Intelligence Industry Convergence Agency, which works on AI industry projects.
- TTA
- The Telecommunications Technology Association, a group that tests and certifies tech products.