Key Points

1. Sustainability of Semiconductor and AI Investment

• The recent controversies and investment directions related to AI and semiconductors have led to the conclusion that “AI is not over.”
• Semiconductors are still essential as a major infrastructure in the AI industry.
• Major Big Tech companies (Microsoft, Meta, Alphabet, etc.) are still actively expanding their investments (Capex increase).
  In particular, investments in new semiconductor technologies and ways to improve efficiency are ongoing.

2. DeepSeek (Dyptic) Issue

• The core of the controversy is the technical result of achieving relatively high performance at a low cost ($5.7 million).
• However, doubts still exist in the market and industry:
  • Uncertainty about the actual cost incurred (media claims of over $1 billion in investment).
  • Possible unauthorized use of some of OpenAI’s models and data.
• Realistically, the DeepSeek case is reinterpreted as the cost of building AI systems is inevitably high, and the effectiveness is still debatable.

3. AI Investment Direction and Monetization

• Transition from the initial “investment-focused” phase to an “efficiency and monetization” phase:
  • The “invest first, monetize later” model has become central to the AI market.
  • Big Tech companies, having invested heavily in AI, are now seeking to transform their business models to generate revenue.
• Important AI monetization methods in the future:
  • B2B: IT solutions, cloud platform services.
  • B2C: Expected to transition to a consumer subscription economy, generating additional revenue based on AI features.

4. NVIDIA and Expansion of Platform Approach

• NVIDIA is no longer limited to hardware but is transitioning to an “AI software platform.”
• Jensen Huang’s strategic approach: Continuously increasing the use of semiconductors through the Cosmos open-source project and the expansion of physical AI such as robots.
• Besides NVIDIA, semiconductor companies also face major challenges in efficiency and optimization.

5. Semiconductor Supply and Demand Outlook:

• Semiconductor demand is still rising across the entire data processing process, from training to inference.
• In particular, the high computational requirements at the inference stage emphasize the sustainability of semiconductor investment.
• Advancements and efficiencies in semiconductor manufacturing technology are also underway, according to Moore’s Law.

6. Key Points to Watch:

• NVIDIA’s announcement of next-generation chip products (e.g., B300).
• Heat generation and technical stability of Luminar and other related components.
• The direction of U.S. technological monopoly and Trump’s policies.
• Market consensus on the speed of AI capital investment; short-term adjustments are a major variable.

< Summary >

1. Semiconductors and AI are closely related, and AI investment and monetization are expected to continuously increase semiconductor demand.
2. The main investment direction is shifting to expanding B2B solutions and B2C subscription-based models.
3. Technological innovations from major semiconductor companies, including NVIDIA, are strengthening the use of software as a platform.
4. Even if semiconductor infrastructure investment slows down in the short term, the medium- to long-term growth potential of the AI and ICT industries is positive.

[More…]

• “NVIDIA Strategy and the Future of AI”
  Link: https://nextgeninsight.net/?s=%EC%97%94%EB%B9%84%EB%94%94%EC%95%84
• “Semiconductor Investment and the Potential Rise of AI”
  Link: https://nextgeninsight.net/?s=%EB%B0%98%EB%8F%84%EC%B2%B4

*YouTube Source: [이효석아카데미]

– 하루가 멀다하고 폭풍이 몰아치는 AI, 반도체 시장에서 살아남을 주식은 바로 ‘이것’ㅣ최도연 SK증권 센터장 [1부]