AMD to establish AI research center in Korea this year, hire hundreds

AMD announced Wednesday that it plans to establish an AI research center in Korea by the end of this year and hire hundreds of researchers. The announcement followed a meeting between AMD CEO Lisa Su and Second Vice Minister of Science and ICT Ryu Je-myung in Seoul. AMD also plans to develop AI talent, build a heterogeneous computing reference model combining AMD CPUs and GPUs with Korean NPUs, and expand infrastructure with MangoBoost.

The meeting reviewed AI cooperation agreed upon by the Ministry of Science and ICT and AMD during President Lee Jae Myung’s attendance at the San Francisco AI Summit in July.

Thirteen Korean AI infrastructure companies discussed cooperation with AMD involving NPUs, software, memory, storage, cloud services, ROCm and access to global markets.

The participating companies were FuriosaAI, HyperAccel, DEEPX, Mobilint, MangoBoost, Panmnesia, Dnotitia, Moreh, Nota AI, Lablup, FADU, Elice Group and Upstage.

AMD said it would expand AI computing infrastructure in Korea in partnership with MangoBoost.

The ministry said it would review proposals from the participating Korean companies and develop them into concrete joint projects with AMD.

Published Oct 7, 2026 5:52 pm KST

AMD Chair and CEO Lisa Su speaks during a meeting with Second Vice Minister of Science and ICT Ryu Je-myung in Seoul, Wednesday. Yonhap

U.S. chip giant AMD announced Wednesday that it plans to establish an artificial intelligence (AI) research center in Korea by the end of this year and hire hundreds of researchers as part of broader AI cooperation between the company and Korea.

The announcement came during a meeting between AMD CEO Lisa Su and Second Vice Minister of Science and ICT Ryu Je-myung in Seoul.

AMD also plans to help develop graduate-level AI and AI transformation talent and collaborate with Korea on a reference model for heterogeneous AI computing. The model would combine AMD’s central processing units (CPUs) and graphics processing units (GPUs) with domestically developed neural processing units (NPUs).

The meeting, arranged by the Ministry of Science and ICT, served as an interim review of AI cooperation agreed upon by the ministry and AMD during President Lee Jae Myung’s attendance at the San Francisco AI Summit in July.

AMD also outlined plans to work with Korea’s private sector to create new business opportunities.

At a subsequent roundtable with 13 Korean companies involved in AI infrastructure, company representatives discussed potential areas of cooperation with Su, including integrating AMD’s CPUs and GPUs with Korean AI chips, participating in open software ecosystems, such as ROCm, and helping Korean AI chipmakers enter global markets.

The roundtable brought together 13 companies spanning key areas of AI infrastructure, including NPUs, software, memory, storage and cloud services. The participants were FuriosaAI, HyperAccel, DEEPX, Mobilint, MangoBoost, Panmnesia, Dnotitia, Moreh, Nota AI, Lablup, FADU, Elice Group and Upstage.

AMD later announced it would expand AI computing infrastructure in Korea in partnership with MangoBoost.

Describing Korea as an important partner for AMD, Su expressed her commitment to creating new opportunities for cooperation with leading Korean companies as the country strengthens its AI competitiveness.

Also noting that AMD has expanded connections among technologies and companies through its open ecosystem, she said AMD will work closely with Korean firms with strong technological capabilities to help them connect with the global AI ecosystem.

The ministry said it will closely review the proposals and requests by Korean companies at the roundtable and develop them into concrete joint projects as part of follow-up efforts with AMD.

“We hope this meeting will be a starting point for combining AMD’s global capabilities with Korean companies’ technological expertise to develop new AI technologies and businesses and further strengthen Korea’s AI semiconductor ecosystem,” Ryu said.

AMD周三宣布，计划于今年年底前在韩国建立人工智能研究中心，并招聘数百名研究人员。此前，AMD首席执行官苏姿丰(Lisa Su)与韩国科学技术信息通信部第二副部长柳济明在首尔举行了会晤。AMD还计划培养人工智能人才，构建结合AMD CPU和GPU与韩国NPU的异构计算参考模型，并利用MangoBoost扩展基础设施。

会议回顾了韩国科学技术信息通信部与AMD在李在明总统7月出席旧金山人工智能峰会期间达成的人工智能合作协议。

十三家韩国人工智能基础设施公司与 AMD 探讨了在 NPU、软件、内存、存储、云服务、ROCm 和全球市场准入等方面的合作。

参与的公司有：FuriosaAI、HyperAccel、DEEPX、Mobilint、MangoBoost、Panmnesia、Dnotitia、Moreh、Nota AI、Lablup、FADU、Elice Group 和 Upstage。

AMD表示将与MangoBoost合作，在韩国扩展人工智能计算基础设施。

该部表示将审查参与的韩国公司的提案，并将其发展成为与AMD的具体联合项目。

发布于2026年10月7日下午5:52（韩国标准时间）

AMD董事长兼首席执行官苏姿丰周三在首尔会见韩国科学技术信息通信部第二副部长柳济明时发表讲话。（韩联社）

美国芯片巨头AMD周三宣布，计划在今年年底前在韩国建立一个人工智能（AI）研究中心，并招聘数百名研究人员，作为该公司与韩国之间更广泛的人工智能合作的一部分。

该消息是在 AMD 首席执行官苏姿丰 (Lisa Su) 与韩国科学技术信息通信部第二副部长柳济明 (Ryu Je-myung) 在首尔会面时宣布的。

AMD还计划帮助培养人工智能和人工智能转型领域的硕士研究生人才，并与韩国合作开发异构人工智能计算参考模型。该模型将结合AMD的中央处理器（CPU）和图形处理器（GPU）以及韩国自主研发的神经处理器（NPU）。

此次会议由科学技术信息通信部安排，是对该部与AMD在李在明总统7月出席旧金山人工智能峰会期间达成的AI合作协议进行中期审查。

AMD还概述了与韩国私营部门合作创造新的商业机会的计划。

在随后与 13 家韩国人工智能基础设施公司举行的圆桌会议上，公司代表讨论了与苏姿丰合作的潜在领域，包括将 AMD 的 CPU 和 GPU 与韩国人工智能芯片集成，参与 ROCm 等开放软件生态系统，以及帮助韩国人工智能芯片制造商进入全球市场。

此次圆桌会议汇聚了13家公司，涵盖人工智能基础设施的关键领域，包括神经网络处理单元（NPU）、软件、内存、存储和云服务。参会公司包括FuriosaAI、HyperAccel、DEEPX、Mobilint、MangoBoost、Panmnesia、Dnotitia、Moreh、Nota AI、Lablup、FADU、Elice Group和Upstage。

AMD 随后宣布将与 MangoBoost 合作，在韩国扩展人工智能计算基础设施。

苏姿丰表示，韩国是 AMD 的重要合作伙伴，她致力于与韩国领先企业创造新的合作机会，以增强韩国在人工智能领域的竞争力。

她还指出，AMD 通过其开放的生态系统扩大了技术和公司之间的联系，并表示 AMD 将与拥有强大技术能力的韩国公司密切合作，帮助它们融入全球人工智能生态系统。

该部表示，将仔细审查韩国企业在圆桌会议上提出的建议和要求，并将其发展成为具体的联合项目，作为与 AMD 后续努力的一部分。

柳表示：“我们希望此次会议能够成为将AMD的全球实力与韩国企业的技术专长相结合，开发新的人工智能技术和业务，并进一步加强韩国人工智能半导体生态系统的起点。”