OpenAI, SK Hynix, and Samsung: Powering the AI Revolution with Stargate
The rapid acceleration of artificial intelligence demands a robust and constantly evolving hardware ecosystem. South Korea, with its dominant semiconductor industry, stands at the forefront of this global technological push, particularly through its collaboration with OpenAI on the ambitious Stargate project.
Source: Own illustration
Quick Summary
- Stargate Initiative: OpenAI’s monumental $500 billion project to build AI data centers globally, with South Korea as a pivotal partner.
- Korean Involvement: SK Group and Samsung are collaborating with OpenAI to develop AI data centers in South Korea and supply critical High Bandwidth Memory (HBM) chips.
- HBM Demand: OpenAI requires up to 900,000 DRAM wafers per month, more than double the current global HBM capacity, representing an estimated $72 billion in additional demand for Korean chip manufacturers over four years.
- Production Expansion: Both SK Hynix and Samsung are significantly increasing their HBM production capacities with new facilities and investments.
- Broader Impact: The partnerships extend to integrating AI technologies across various sectors, including next-generation data center solutions and corporate AI adoption.
The Stargate Initiative and South Korea’s Central Role
OpenAI’s Stargate project, a monumental $500 billion undertaking in collaboration with Oracle and SoftBank, aims to construct a network of AI data centers primarily in the United States. For more details on this collaboration, see OpenAI's official announcement regarding Samsung and SK joining Stargate. However, its reach extends globally, with South Korea emerging as a pivotal partner.
On October 1, 2025, SK Group and OpenAI signed an initial agreement in Seoul, signifying their intent to develop an AI data center within the country, as also detailed in the same OpenAI announcement. This collaboration extends beyond data center development, encompassing the critical supply of High Bandwidth Memory (HBM) chips, fundamental components for advanced AI, according to KED Global.
Source: businessinsider.com
Sam Altman, OpenAI’s CEO, has met with key executives from Samsung and SK to secure crucial partnerships for the Stargate initiative, solidifying these important alliances.
OpenAI’s CEO, Sam Altman, has actively engaged with leading South Korean manufacturers, meeting with figures like Samsung Electronics Executive Chairman Jay Y. Lee and SK Chairman Chey Tae-won to solidify these partnerships, as reported by KED Global. These agreements, made after a meeting in Seoul that included South Korean President Lee Jae-myung, underscore the strategic importance of this alliance. The South Korean presidential office indicated that OpenAI intends to form joint ventures with Samsung and SK Hynix to establish two data centers in South Korea.
❝ the nation possesses the vision, technology, talent, and infrastructure to lead globally in AI. ❞
OpenAI CEO
High Bandwidth Memory: The Core of AI Computation
The cornerstone of these agreements lies in the provision of HBM chips. Both Samsung Electronics and SK Hynix, which together control nearly 80 percent of the global HBM market, have committed to supplying the necessary DRAM wafers for the Stargate project, according to KED Global.
Unprecedented Demand
OpenAI’s demand is staggering: up to 900,000 DRAM wafers per month, a quantity exceeding double the current global HBM industry capacity. This immense requirement, equivalent to roughly 40% of the entire global DRAM production, highlights the unprecedented scale of the Stargate initiative. Analysts estimate this commitment represents over 100 trillion Won (approximately $72 billion USD) in additional demand for Korean chip manufacturers over four years.
SK Hynix’s Vision
SK Hynix, a key player, will develop a production system to meet this demand, while also actively supporting OpenAI’s GPU procurement strategy. The company’s vision is to become a "Full-Stack AI Memory Creator," presenting a strategy that includes:
- Custom HBM: Designed to enhance performance by integrating computational functions directly onto the HBM base die, functions typically handled by GPUs or ASICs.
- AI DRAM (AI-D): Encompasses optimized DRAM products such as MRDIMM, SOCAMM2, LPDDR5R, CXL Memory Module (CMM), and Processing-in-Memory (PIM).
- AI NAND (AI-N): Catering to future AI systems, offers high performance, bandwidth, and ultra-high capacity.
SK Hynix showcased its 12-layer HBM4 and NVIDIA’s GB300 Grace Blackwell Superchip with SK Hynix's 36 GB HBM3E at the SK AI Summit 2025.
Source: news.skhynix.co.kr
At the SK AI Summit 2025, SK Hynix unveiled its advanced 12-layer HBM4 chip, showcasing its leadership in next-generation memory solutions for AI.
Samsung’s Contributions
Samsung also plans to scale its production of High-Bandwidth DRAM memory chips for Stargate and the new AI data centers. Beyond HBM, Samsung will likely supply Graphics DRAM (GDDR), high-capacity SSDs, and Low-Power DRAM with In-Memory Computing (LPDDR5X-PIM) tailored for AI applications. The company’s Samsung SDS Co Ltd subsidiary has already signed a contract with OpenAI for the construction and operation of data centers under the Stargate umbrella. Further strategic ventures include Samsung C&T and Samsung Heavy Industries exploring the development of floating offshore data centers.
Expansion and Investment in Production Capacity
Meeting OpenAI’s extraordinary demand necessitates significant expansion in production capabilities from both Samsung and SK Hynix.
SK Hynix’s Expansion
SK Hynix’s current HBM capacity stands at approximately 160,000 wafers per month. To bridge this gap, the company announced plans to increase infrastructure investments by more than four times its previously stated figures. New facilities like the Cheongju M15X plant, expected to complete by late 2025, and Phase 1 of the Yongin Semiconductor Cluster, slated for May 2027, are central to this strategy.
Samsung’s Expansion
Samsung, for its part, aims to expand its HBM production capacity by about 50 percent in 2026. Its P5 facility in Pyeongtaek, Gyeonggi, is projected to be operational by 2028.
Source: businesskorea.co.kr
Samsung’s P5 facility in Pyeongtaek, Gyeonggi, is a major investment in expanding its HBM production capabilities for future AI demands.
Market Impact
The increased demand for HBM has already impacted the market, with Samsung and SK Hynix raising memory chip prices by up to 30 percent in the fourth quarter of 2025. The intensified focus on high-value HBM could also lead to a reallocation of DRAM production, potentially tightening the supply of mainstream DRAM and driving up prices.
The Broader AI Ecosystem Impact
The Stargate project and its associated partnerships extend beyond HBM, signaling a comprehensive integration of AI technologies across various sectors. SK Telecom will collaborate with OpenAI on the development, construction, and operation of the AI center in South Korea, fostering solutions for next-generation data centers and exploring diverse AI service applications for consumers and businesses. SK Group and Samsung will also integrate ChatGPT Enterprise and OpenAI APIs into their operations, signifying a broader adoption of AI tools within their corporate structures.
Table: HBM Market Share Q3 2025
| Company | HBM Market Share (Q3 2025) |
|---|---|
| SK Hynix | 53% |
| Samsung | 35% |
| Micron Technology Inc. | In Development (HBM4 for 2026) |
OpenAI and AMD Partnership
OpenAI’s partnership with AMD also contributes to the evolving AI hardware landscape. OpenAI will deploy 6 gigawatts of AMD’s Instinct graphics processors over several years, starting with a 1-gigawatt rollout in the latter half of 2026. This multi-billion dollar deal, estimated between $60 billion and $90 billion, underlines the diverse needs of AI infrastructure. While SK Hynix held a 53% market share in HBM in Q3 2025, compared to Samsung’s 35% (according to Counterpoint Research), Samsung Electronics may hold an advantage in future HBM4 chip orders due to its closer ties with AMD. Micron Technology Inc. is also gearing up for mass production of HBM4 memory chips in 2026 for next-generation AI platforms.
Frequently Asked Questions
What is OpenAI's Stargate project?
The Stargate project is a massive $500 billion initiative by OpenAI, in collaboration with partners like Oracle and SoftBank, to build a network of AI data centers, primarily in the United States, to support the immense computational needs of advanced AI models.
How much HBM will OpenAI need?
OpenAI's demand for HBM chips is projected to be up to 900,000 DRAM wafers per month. This is more than double the current global HBM industry capacity and could account for up to 40% of the entire global DRAM production.
What is the role of South Korean companies in Stargate?
South Korean giants Samsung and SK Group are pivotal partners. They are committed to supplying the vast quantities of HBM chips needed for Stargate and are also collaborating with OpenAI to develop and operate AI data centers in South Korea.
What are HBM chips and why are they important for AI?
HBM (High Bandwidth Memory) chips are a type of high-performance RAM that provides significantly higher bandwidth than traditional DRAM. They are crucial for AI applications because they can quickly transfer the massive amounts of data required for training and running complex AI models, like large language models.
How will this impact the semiconductor market?
The unprecedented demand for HBM is leading to significant expansion in production capacities by Samsung and SK Hynix. It has already resulted in memory chip price increases and may cause a reallocation of DRAM production, potentially affecting the supply and pricing of mainstream DRAM.
Conclusion
The collaboration between OpenAI, Samsung, and SK Group marks a significant chapter in the global development of artificial intelligence. By funneling massive investments into South Korea’s semiconductor manufacturers and leveraging their unparalleled expertise in HBM technology, the Stargate initiative is not merely building data centers; it is establishing the physical bedrock for the AI era. The scale of the demanded HBM chips, the rapid expansion of production facilities, and the strategic integration of AI across corporate operations collectively signify a profound shift. South Korea is not just a supplier; it is an indispensable partner in shaping the future of AI.