🚨 Stargate Project

Today Korean Social News for Beginners | 2025.11.22

0️⃣ US $450B AI Infrastructure Investment and Korea's Response Strategy

📌 US Invests $450B in AI Infrastructure 'Stargate'... Korea Urgently Needs Response to Avoid Tech Dependency

💬 The United States is launching the 'Stargate' project, a massive artificial intelligence infrastructure initiative with OpenAI, SoftBank, Oracle, and MGX. This plan will invest over $450 billion by 2028 to build five 10GW-scale AI data centers, and experts see it as a key strategy that will change the global AI competition. The project name 'Stargate' means AI technology is a gateway leading humanity to a new era, and it is a strategic investment to strengthen America's technology leadership. Korea has launched a 150 trillion won National Growth Fund in response, but experts point out that much larger investments and infrastructure expansion are urgently needed to become one of the 'top 3 AI powers' rather than an AI-dependent country. Beyond technology and talent competition, securing capital and energy has become the new standard for national competitiveness.

💡 Summary

• Stargate is a massive $450 billion AI infrastructure project led by the United States.
• By 2028, it will build five 10GW-scale AI data centers to secure global AI market leadership.
• Korea needs comprehensive responses including capital, energy, and infrastructure to escape AI dependency.

1️⃣ Definition

The Stargate Project is a massive artificial intelligence infrastructure initiative that the United States is pursuing with OpenAI, Oracle, SoftBank, and Microsoft-affiliated MGX, investing over $450 billion by 2028 to build large-scale data centers and supercomputing facilities.

The core of this project is to build large-scale computing infrastructure for AI training and applications. The goal is not simply technology development, but securing the physical foundation to train and operate AI models - data centers and power infrastructure. The project name 'Stargate' comes from science fiction movies meaning a gateway to another dimension, symbolizing that AI is a gateway leading humanity to a new era.

💡 Why Is This Important?

• AI competition is more about infrastructure and power to train models than algorithms.
• It is a strategic investment that determines global AI market leadership.
• If Korea cannot build an independent AI ecosystem, technology dependency may deepen.
• It becomes a new standard for national competitiveness covering semiconductors, power, and cloud.

2️⃣ Current Status of Stargate Project and Global AI Competition

📕 Scale and Goals of the Stargate Project

• It is a massive project requiring huge capital and energy. Key details include:

  • Investing over $450 billion by 2028.
  • Building five 10GW-scale AI data centers, equivalent to the total power production of a small country.
  • Global companies participate including OpenAI (ChatGPT developer), Oracle (cloud infrastructure), SoftBank (investment), and MGX (Microsoft-affiliated investment firm).
  • The first data center will be built in Texas and gradually expanded.

• It is a strategic investment to secure technology leadership. Main objectives include:

  • Dramatically increasing US AI data processing capabilities to secure global market leadership.
  • A national-level strategic project to gain advantage in AI technology competition with China.
  • Integrating semiconductors, power, and cloud infrastructure to dominate the entire AI ecosystem.
  • Establishing a foundation for US companies to safely develop and operate AI models domestically.

📕 Changes in Global AI Infrastructure Competition

• The core of AI competition is shifting to infrastructure. Major changes include:

  • In the past, excellent algorithms and AI model development were the core of competition.
  • Now, securing computing power and electricity to train large-scale AI models has become more important.
  • Training massive AI models like ChatGPT requires hundreds of billions of won and enormous power.
  • Capital power for data center construction and operation has become what determines national competitiveness.

• Energy security has emerged as a new competitive factor. Key issues include:

  • A 10GW-scale data center equals the power consumption of about 1 million households.
  • The US can supply large-scale power through natural gas and nuclear power, but Korea faces power supply difficulties.
  • Environmental issues must also be considered, requiring a combination of renewable energy and nuclear power.
  • Power securing capability has become a key factor determining whether a country becomes an AI power.

📕 Korea's Current Response

• Korea launched a 150 trillion won National Growth Fund. Key details include:

  • The Korean government created the National Growth Fund for AI semiconductor and data center investment.
  • Planning total investment of 150 trillion won including private capital.
  • Semiconductor companies like SK Hynix and Samsung Electronics are actively investing in AI chip development.
  • IT companies like Naver and Kakao are also developing their own AI models.

• The assessment is that scale and speed are still insufficient. Main limitations include:

  • Compared to US Stargate ($450 billion), the investment scale is one-third.
  • Power supply for data center construction is the biggest obstacle.
  • AI talent training and recruitment lag behind the US and China.
  • Without building an independent AI ecosystem, there is risk of becoming dependent on US or Chinese technology.

💡 Key Issues of the Stargate Project

1. Capital Competition: $450 billion investment determines national-level competitiveness
2. Energy Security: 10GW power supply capacity is key to AI infrastructure construction
3. Technology Dependency: Risk of US-China dependency without independent ecosystem
4. Industry Reorganization: AI infrastructure integrates semiconductor, cloud, and power industries
5. National Security: Era where AI technology leadership determines economy and security

3️⃣ Korea's Strategy to Become an AI Powerhouse

✅ Expanding Large-Scale Investment and Securing Capital

• Investment scale must be significantly expanded. Key directions include:

  • Considering plans to expand current 150 trillion won investment to over 300 trillion won.
  • Utilizing various funding sources including government budget, private capital, and national pension.
  • Inducing corporate AI investment through long-term low-interest loans and tax benefits.
  • Diversifying capital procurement channels by attracting global investors.

• Strategic investment priorities must be set. Key tasks include:

  • Concentrating investment on AI semiconductors, data centers, and power infrastructure.
  • Building strengths in industry-specific AI rather than general-purpose AI.
  • Focusing on AI solution development in areas where Korea has competitiveness like healthcare, manufacturing, and finance.
  • Expanding programs supporting AI utilization by small businesses and startups.

✅ Expanding Energy Infrastructure

• Large-scale power supply system must be built. Key measures include:

  • Reviewing construction of dedicated power grid for AI data centers.
  • Actively promoting nuclear power expansion and small modular reactor (SMR) introduction.
  • Increasing renewable energy share while expanding energy storage systems for stable power supply.
  • Developing data center design technology with high power efficiency.

• Eco-friendly energy transition must be parallel. Key tasks include:

  • Investing in technology to reduce carbon emissions from data centers.
  • Seeking energy recycling methods like district heating using waste heat.
  • Finding balance between responding to international environmental regulations and maintaining AI industry competitiveness.
  • Considering overseas data center construction to overcome energy constraints.

✅ Talent Development and Ecosystem Creation

• AI professional workforce must be significantly expanded. Key directions include:

  • Expanding university AI-related departments and enrollment quotas, strengthening practical education.
  • Visa system improvement and better treatment needed to attract excellent overseas talent.
  • Expanding AI retraining programs for current employees.
  • Activating talent exchange through cooperation with global AI companies.

• AI ecosystem must be activated. Key tasks include:

  • Building networks where large companies, SMEs, startups, universities, and research institutes cooperate.
  • Establishing data sharing platforms for AI model development.
  • Expanding regulatory sandboxes to support innovative AI service launches.
  • Strengthening international cooperation to actively participate in global AI ecosystem.

4️⃣ Related Terms Explanation

🔎 AI Data Center

• An AI data center is a massive computing facility for artificial intelligence model training and operations.
  • An AI data center is a facility concentrating high-performance servers, storage devices, and network equipment needed to train and run artificial intelligence models. Unlike general data centers, they have very high parallel computing capability by mounting large quantities of GPUs (Graphics Processing Units) or AI-dedicated chips.
  • Characteristics of AI data centers include: First, they consume enormous power. A 10GW facility equals the power consumption of about 1 million households. Second, cooling systems are very important. Heat generated by high-performance chips must be efficiently cooled for stable operation. Third, ultra-high-speed networks are essential. Thousands or tens of thousands of chips must work together simultaneously, requiring fast data transmission. Fourth, physical security and cybersecurity are very important.
  • The Stargate project builds five such AI data centers, with each center equipped with hundreds of thousands of GPUs and capable of simultaneously training multiple massive AI models like ChatGPT. For Korea to build such facilities, various challenges must be solved including power security, site selection, and cooling technology.

🔎 AI Semiconductors (GPU, NPU, TPU)

• AI semiconductors are high-performance chips specialized for artificial intelligence operations.
  • AI semiconductors are semiconductors that efficiently process massive parallel operations needed for training and inference of artificial intelligence models. Traditional CPUs are strong at sequential processing, but AI requires performing millions of calculations simultaneously, so parallel processing chips like GPUs are needed.
  • Main AI semiconductor types include: First, GPUs (Graphics Processing Units) were originally for graphics processing but are very effective for AI training. NVIDIA is the leading company. Second, TPUs (Tensor Processing Units) are AI-dedicated chips developed by Google, optimized for tensor operations. Third, NPUs (Neural Processing Units) are chips specialized for neural network operations, also mounted on smartphones and edge devices.
  • Korea's Samsung Electronics and SK Hynix are actively investing in AI semiconductor development. Especially in HBM (High Bandwidth Memory), they have world-class technology, so there is high possibility of gaining competitiveness in the AI semiconductor market. However, more investment and technology development are needed to catch up with leading companies like NVIDIA.

🔎 Massive AI Models

• Massive AI models are large-scale artificial intelligence with hundreds of millions to trillions of parameters.
  • Massive AI models are large-scale artificial intelligence models trained with hundreds of millions to trillions of parameters based on deep learning technology. ChatGPT's GPT-4 is known to have about 1 trillion or more parameters, and training it costs hundreds of billions of won and takes several months.
  • Characteristics of massive AI include: First, they train with vast amounts of data. They utilize almost all public data from the internet including text, images, and videos. Second, they have high versatility. They can perform various tasks like translation, summarization, coding, and creation, not just one task. Third, they require enormous computing resources. Training alone requires running tens of thousands of GPUs for several months.
  • The Stargate project builds infrastructure to safely train and operate such massive AI models domestically. Korea is also developing its own massive AI like Naver's HyperCLOVA X and Kakao's KoGPT, but there is still a gap with global standards, requiring more investment, data, and infrastructure.

🔎 AI Technology Leadership

• AI technology leadership means power to dominate economy and security through artificial intelligence technology.
  • AI technology leadership means securing dominance in global economy, military, and politics by dominating artificial intelligence technology and infrastructure. Just as steel and coal determined national power during the Industrial Revolution, AI technology has become a key factor determining national competitiveness in modern times.
  • Importance of AI technology leadership includes: First, enormous economic value. AI is applied to all industries, dramatically increasing productivity. Second, it brings military superiority. AI plays a decisive role in weapon systems, information analysis, and cyber warfare. Third, it strengthens information control. AI can analyze data, understand public opinion, and make policy decisions. Fourth, it has authority to set standards. Leading countries lead AI ethics, regulations, and technology standards.
  • Currently, the US and China are competing for AI technology leadership. The US responds with public-private cooperation projects like Stargate, while China responds with state-led investment. Korea must build its own AI ecosystem, strengthen cooperation with allies, and build competitiveness in specialized areas to not fall behind in this competition. AI technology dependency means economic and political dependency, so national-level strategic response is needed.

5️⃣ Frequently Asked Questions (FAQ)

Q: How will the Stargate Project affect the Korean economy?

A: Positively, increased semiconductor exports are expected, but negatively, there are concerns about widening AI technology gaps.

• The Stargate project requires massive amounts of AI semiconductors, so it's an opportunity for Korean semiconductor companies. HBM (High Bandwidth Memory) and AI semiconductors produced by Samsung Electronics and SK Hynix are likely to be supplied in large quantities to US data centers. In the short term, increased exports and a semiconductor industry boom are expected.
• However, there are long-term concerns. If the US concentrates AI infrastructure domestically, Korean companies may become dependent on US platforms. For example, to develop services like ChatGPT, Korean companies would need to rent US data centers, potentially losing technology and data sovereignty. Also, because the US investment scale is so large, it may become difficult for Korea to build an independent AI ecosystem. Therefore, rather than being satisfied with short-term profits from semiconductor exports, more investment must be made in building our own AI infrastructure and platforms.

Q: What does the Stargate Project mean for ordinary citizens?

A: AI services will advance further, but dependency on US technology may increase.

• Due to Stargate, AI services like ChatGPT will become more advanced and diverse. People will be able to use faster and more accurate AI assistants, medical diagnostics, education, and creative tools. Korean companies will also launch new products and services using these AI technologies. Daily life will become more convenient and work productivity will increase.
• However, there are negative aspects. If most AI services depend on US companies, personal information and data will flow to the US, potentially causing privacy problems. Also, AI technology gaps may intensify job polarization. Income gaps may widen between high-skilled workers who use AI well and those who don't. At the national level, AI education must be strengthened and policies prepared so all citizens can benefit from AI technology.

Q: What is most important for Korea to become an AI powerhouse?

A: Power security, large-scale investment, and talent development must happen simultaneously.

• There are three core tasks for Korea to become an AI powerhouse. First, power security. AI data centers consume enormous power, so stable and sufficient power must be supplied by expanding nuclear and renewable energy. Without solving the power shortage problem, data centers cannot be operated no matter how much capital and technology there is. Second, large-scale investment. Current investment of 150 trillion won is insufficient; it must be expanded to at least 300 trillion won. All funding sources including government budget, private capital, and national pension must be mobilized.
• Third, talent development. Since AI technology is ultimately developed and operated by people, world-class AI researchers and engineers must be trained and recruited. University education must be strengthened, global talent actively recruited, and employee retraining expanded. When these three are promoted harmoniously, Korea can rise as an AI powerhouse alongside the US and China. If any one is lacking, there is risk of falling into technology dependency, so a total national response is needed.

Table of Contents

• Go to Guide