With the sudden resignation of Nvidia CEO Jensen Huang, the domestic artificial intelligence (AI) industry is watching the impact on cooperation in the physical AI field such as graphics processing unit (GPU) supply, large-scale AI infrastructure investment, and autonomous driving and robots.
As NVIDIA has announced that it will maintain existing contracts and product development plans, the general view in the industry is that there is not a high possibility that there will be immediate disruption in supply or business.
However, there are observations that the schedule and priorities for next-generation GPU volume allocation, additional investment, and technical support may vary depending on the judgment of the new management.
◇ “Limited disruption to existing GPU supply and cooperation business”
According to the information and communications technology (ICT) industry on the 13th, Nvidia announced on the 12th (local time) that CEO Hwang had immediately resigned from his position as president, CEO, and board member. Chief Financial Officer (CFO) Colette Kress took on the role of interim CEO.
NVIDIA has stated that it will keep its promises to existing customers and implement its product development roadmap as planned despite the change in management.
The domestic industry also believes that already confirmed GPU supply contracts and ongoing cooperation projects will not be significantly affected.
The domestic GPU introduction plan announced during CEO Hwang's visit to Korea last year totals more than 260,000 units. Samsung Electronics [005930], SK Group, and Hyundai Motor Group each secure 50,000 copies, Naver Cloud secures 60,000 copies, and the government secures more than 50,000 copies.
First of all, the government believes that there will be no disruption in the supply of existing contracted quantities.
Based on the NVIDIA B200 GPU, about 13,100 GPUs secured through last year's supplementary budget have already been built and are being used, and about 8,500 GPUs for the 6th National Supercomputer have also been built. In addition, it is explained that the introduction of the latest GPUs, such as Vera Rubin, which will be additionally carried out this year, will proceed as planned.
Companies are in a similar mood.
An official from a company in cooperation with NVIDIA said, "We are communicating with the field regarding the ongoing business, but the feeling is that there is not a high possibility of problems occurring immediately." He added, "Already documented supply and investment contracts are not easy to be shaken by a change in management."
In particular, since the AI data center business is a long-term project that is promoted over 10 to 15 years from equipment introduction to financing, facility construction, capacity expansion, and customer acquisition, it is analyzed that the possibility of existing contracts being shaken just by changing management is limited.
The fact that Korea and Nvidia have a close interdependent relationship in the semiconductor supply chain is also considered a factor supporting supply stability. This is because domestic companies are playing a key role in the production of NVIDIA's AI accelerators, such as Samsung Electronics and SK Hynix [000660] supplying high bandwidth memory (HBM) to NVIDIA.
◇ Next-generation GPU uncertainty… Cooperation follow-up attention
However, there is a view that the speed of large-scale new investments or strategic cooperation may change in the process of reorganizing NVIDIA's decision-making system, where major management decisions were focused on CEO Jensen Huang.
A domestic industry insider pointed out, "If NVIDIA readjusts its supply priorities, the biggest concern is that it may affect the introduction schedule of 'Vera Rubin,' which does not have a large quantity," and added, "AI startups and research institutes that lack the capacity to secure their own GPUs are also bound to be sensitive to changes in supply policy."
Attention is also focused on the strategic cooperation between NVIDIA and domestic companies that has expanded beyond simple GPU supply to AI data centers, AI factories, and physical AI.
As a representative example, Naver, which recently attracted equity investment from NVIDIA, is one of the companies receiving attention for its future AI infrastructure expansion and business direction.
Naver signed a contract to secure computing infrastructure worth up to $9 billion with Brookfield, a global asset management company, last month, and received equity investment of $1 billion (approximately KRW 1.3412 trillion) from Nvidia on the 6th. Nvidia secured approximately 4.13% of Naver's shares through a paid-in capital increase allocated to a third party, becoming the third largest shareholder.
To build a 1 gigawatt (GW) global AI factory, Naver Cloud operates the business, Brookfield invests funds, and NVIDIA is responsible for GPU supply and technical cooperation.
Naver plans to operate a 55-megawatt (MW) AI factory in the first half of 2027 based on Sejong City's data center 'Gak Sejong' and expand to 200 MW by 2028.
It is expected that there will be no major problems with business continuity as equity investment and infrastructure securing contracts have already been concluded, but considering future infrastructure expansion and introduction of next-generation GPUs, there is analysis that the new management's judgment may be affected in terms of additional investment and technical support.
In addition to Naver, NVIDIA has recently begun collaborating with domestic companies one after another and has been concretizing new businesses in the AI infrastructure and physical AI fields.
It announced cooperation with the Ministry of Science and ICT to create a physical AI ecosystem, and signed a business agreement with SK Telecom and Rebellion with the goal of building a heterogeneous AI computing system combining GPU and neural network processing unit (NPU) and applying it to AI data centers.
It is also pursuing cooperation with Kakao Mobility in the autonomous driving and mobility AI fields, and has also announced investment plans for domestic AI startups Upstage, Motif Technologies, and Real World.
However, as such technical cooperation and joint development projects require continuous coordination of development schedules, scope of technical support, and additional investment scale even after the initial agreement, it is pointed out that the actual speed of commercialization may vary depending on subsequent decisions.
In particular, the fact that NVIDIA's major management decisions have been focused on CEO Hwang is pointed out as a variable.
Another industry official said, "Nvidia has decided on major issues centered on a small number of executives who report directly to CEO Hwang rather than fully delegating decision-making authority to major business divisions, so I think decision-making may be somewhat delayed after the management change."
However, to date, no specific circumstances have been confirmed that NVIDIA is attempting to withdraw or reduce cooperation or investment plans announced with domestic companies.
◇ Hyundai Motors’ autonomous driving and AI factory cooperation is also keen
Hyundai Motor Group, which has maintained a special cooperative relationship with Nvidia through the 'Kanbu Meeting', is also paying close attention to the impact of CEO Hwang's resignation.
As the company has been expanding its strategic partnership with Nvidia across future businesses, including autonomous driving, robotics, and AI infrastructure, it appears to be paying close attention to the impact the replacement of Nvidia's head will have on its future business strategy.
Hyundai Motor Group signed a strategic partnership with NVIDIA for mobility innovation in January last year, and in October of the same year, decided to join forces in developing the domestic physical AI ecosystem, including building an AI factory based on NVIDIA's next-generation AI chip 'NVIDIA Blackwell'.
Cooperation is also underway in the autonomous driving field, including developing next-generation autonomous driving technology using NVIDIA's autonomous driving platform. The 2028 self-driving car mass production roadmap presented by Hyundai Motor Company and Kia Motors is also based on cooperation with NVIDIA.
Earlier this year, the company hired CEO Park Min-woo, an autonomous driving technology expert from NVIDIA, as head of the group's Advanced Vehicle Platform (AVP) and CEO of 42dot.
Inside and outside the Hyundai Motor Group, it is expected that existing businesses and projects will proceed as planned as the two companies already have an important cooperative relationship.
However, as the business conditions and implementation plan have not yet been specifically formalized, such as the content of CEO Hwang's meeting with Hyundai Motor Group Chairman Chung Eui-sun to discuss cooperation, the variable is whether the new management will continue existing discussions in the future.
CEO Hwang met with Chairman Chung last June and received a proposal to build a Saemangeum data center, saying, "I responded that I would do it if there was 'good barbecue pork'." In July, he also told President Lee Jae-myung, who visited the United States, that "we will jointly develop autonomous driving Genesis vehicles through a partnership with Hyundai Motor Group."
An industry official said, "The fact that Nvidia's cooperation with domestic companies was promoted with speed is not unrelated to CEO Hwang's unique style. Since his successor is a financial expert and has a cautious style, there are concerns that the speed of business promotion may not be as fast as before or priorities may change."
