Microsoft CEO to meet with Samsung, SK Hynix, LG, SK Telecom chiefs

Microsoft Corp.’s top executives will meet with leaders of South Korea’s four top technology companies next month to discuss ways to strengthen their partnerships in the artificial intelligence business.

Kyung Kye-hyun, head of Samsung Electronics Co.'s semiconductor business, Kwak Noh-jung, chief executive of SK Hynix Inc., Cho Joo-wan, CEO of LG Electronics Inc. and SK Telecom Co. CEO Ryu Young-sang were invited to the MS CEO Summit 2024, industry sources said on Monday.

At the three-day summit meeting on May 14 in Redmond, Washington, Microsoft Chairman and CEO Satya Nadella and founder Bill Gates will hold closed-door meetings with the Korean tech firms' leaders to strengthen their AI partnerships, sources said.

Microsoft is expected to explore opportunities for joint development and supply of AI chips with Samsung and SK Hynix – the world’s two largest memory chipmakers – and cooperate in AI technology for TVs and other home appliances with LG Electronics. With SK Telecom, the US tech giant is expected to discuss business tie-ups in the 5G telecommunications and cloud sectors, sources said.

“It is unusual for MS to invite the Korean Big Tech companies to participate in such a summit and separate meetings,” said a senior official at a Korean tech company.

WIN-WIN 

The planned gathering between leaders of Microsoft and the Korean tech firms comes as chipmakers, electronics makers and telecommunications companies seek business tie-ups to gain the upper hand in the booming AI business.

Analysts said the Microsoft chiefs’ talks with the Korean CEOs will be mutually beneficial.

If Microsoft installs its AI services on Samsung and LG’s electronics goods, it will boost the US firm’s status in competition with rivals such as Google and Meta Platforms, industry officials said.

Samsung produces about 300 million smartphones annually while LG Electronics sells 100 million home appliance products a year.

Microsoft is already a key customer of Samsung and LG’s AI-enabled smartphones, TVs and other electronic devices.

The US company is also a big buyer in the server memory chip market.

Samsung and SK Hynix will likely propose ways to make next-generation high-bandwidth memory (HBM) AI chips, Compute Express Link (CXL) devices and solid-state drives (SSDs) for servers, all custom-tailored for Microsoft, sources said.

MACH-1, 2 AI CHIPS

Kyung, chief executive of Samsung’s Device Solutions (DS) division, which oversees its chip business, said last month the company is developing AI accelerators, dubbed Mach-1 and Mach-2, and it will begin mass production of a prototype by the end of this year.

Samsung has already agreed to supply Mach-1 AI chips in a deal worth up to 1 trillion won ($752 million) to Korea’s Naver Corp. by year-end, people familiar with the matter said last month.

With the contract, Naver plans to significantly reduce its reliance on Nvidia Corp. for chips used in AI inference.

Mach-1 is an AI accelerator in the form of a system-on-chip (SoC) that reduces the bottleneck between the graphics processing unit (GPU) and HBM chips, according to Samsung.

Leveraging its sale of Mach-1 chips to Naver, Samsung plans to expand its client base to Big Tech firms. Samsung is already in supply talks with Microsoft and Meta Platforms, sources said.

SK TELECOM LEADS TELCO AI ALLIANCE

SK Telecom, Korea’s top mobile carrier, is leading a group of global telecommunications companies in the AI sector, jointly launched with Deutsche Telekom, Singapore Telecommunications Ltd. (SingTel) and United Arab Emirates' e&.

Japan’s SoftBank Group joined the group, the Global Telco AI Alliance (GTAA), earlier this year.

The GTAA members have a combined 1.3 billion users in the world’s key markets, including the US, Europe, the Middle East, Asia, Africa and Australia.

SK Telecom CEO Ryu is expected to discuss ways to cooperate in building AI infrastructure and large language models (LLMs) for generative AI services.

Write to Jeong-Soo Hwang at hjs@hankyung.com


In-Soo Nam edited this article.