KR — 전자신문

Rep. Im Moon-young urges AMD to set up AI research center in Gwangju, citing five location strengths

ChipsNovember 3, 20262 min read

Lim Mun-young, a Korean lawmaker, is asking AMD to build an AI research center in Gwangju. He met AMD's senior vice president on the 3rd and listed five reasons why Gwangju is a good place for it.

A lawmaker pushes for an AMD center in Gwangju

Lim Mun-young, a member of the ruling Democratic Party, wants AMD to set up an AI Center of Excellence (CoE) in Gwangju. On the 3rd, he met Keith Strier, AMD's senior vice president, at the National Assembly in Seoul. He said the National Assembly will support the plan if AMD builds the center there.

Key facts and numbers

Five strengths of Gwangju were listed: a planned AI data center using Korean-made NPU chips, investment support in an R&D zone, a planned chip factory, advanced optical chip research, and closeness to the Seoul area.
The NPU data center in Gwangju is planned to start running in 2030.
A Seoul-to-Gwangju chip cluster trip by KTX (high-speed train) takes about 2 hours.
The lawmaker first proposed the idea to AMD CEO Lisa Su on the 7th of last month, in a personal letter.

Why this talk happened

This meeting is a follow-up to the earlier proposal to AMD's CEO. The lawmaker also suggested joint research and technology tests between AMD and Korean AI chip companies and research groups. Keith Strier leads AMD's global AI strategy and its work with government and private groups. He once held a similar AI role at Nvidia, where he led national AI projects.

Why it matters

A research center could bring jobs, research money, and AI skills to the Gwangju region. It also shows how Korea wants to become a bigger player in AI chips. Cheap clean power from renewable energy is another selling point for AI data centers, which use a lot of electricity.

What to watch

Whether AMD responds to the proposal and agrees to set up a center in Gwangju.
Whether the National Assembly and local government put real money or rules behind their support.
Progress on the planned NPU data center, which is set to start in 2030.
Any news on joint research between AMD and Korean AI chip companies.

Glossary

AMD: A U.S. company that makes computer chips, including chips used for AI.
KTX: Korea's high-speed train network.
Nvidia: A U.S. company that makes AI chips and graphics chips.

FPT Begins Trial Operations of Chip Plant in Bac Ninh

FPT has started trial operations at its chip testing and packaging plant in Bac Ninh, which has six production lines and a maximum design capacity of 300 million chips per year. The 1,600-square-meter facility is located in Yen Phong II-C industrial park, and the company plans to fulfill its first chip testing order for a South Korean partner. Initial work focuses on functional testing and reliability checks, and the plant will also serve as a training site for students from FPT Polytechnic starting in December.

ChipsThe RegisterNov 3

Toyota shows off SoC that stores model weights and its OS on integrated memory

Toyota major shareholder AISIN has developed a system-on-chip built on a 16nm FinFET process with an Arm Cortex-A53 dual-core CPU and Arm Mali-G31 GPU, using built-in STT-MRAM to store AI model weights and an operating system. AISIN and Tohoku University say the chip reaches kernel startup in 65.7 milliseconds, 38 times faster than using external memory, and can run the Tiny YOLOv3 object detection algorithm. AISIN has not said if or when the chip will enter production.

ChipsThe RegisterNov 3

Marvell wants to be a copper topper, bets $115B on photonics leader to push optical networking

Marvell Technology plans to acquire photonics company Lumentum in an all-stock deal valued at $115 billion, expected to close in the first half of 2028. The move aims to fill gaps in Marvell's photonics portfolio as AI infrastructure builders such as Nvidia and AMD pursue larger rack-scale systems that rely on optical interconnects. The article notes that copper reach falls sharply as per-lane speeds rise, while optics can cover kilometers and reduce rack power and cooling needs.

ChipsVnExpressNov 3

AI Chips Attract Chinese Synthetic Diamond Producers

Many Chinese synthetic diamond manufacturers are shifting toward AI chips and semiconductors in pursuit of higher profits than the jewelry market offers. Facing intense competition and price pressure, these producers, who account for about half of global supply, are using diamond for heat dissipation and other advanced applications. Analysts say diamond's far higher thermal conductivity than copper could make it valuable for cooling AI data-center hardware.