Summary of Key Points
Recently, there has been a trend of CPU shortages and price increases. The market predicts that by 2026, Intel chips will be more in demand than NVIDIA chips, leading to an eightfold surge in Intel's stock price. The underlying reason for this is that the AI industry is transitioning from the training phase, where models are taught how to learn, to the application phase, where models are used to perform tasks directly for people. With the advent of "agent-based AI systems" that can complete tasks independently, the demand for CPUs will increase significantly, and Intel's strengths in this field will become even more evident.
Detailed Analysis
1. Why is there a sudden shortage and price increase of CPUs? —— The explosion of AI applications has overwhelmed the supply chain
Previously, AI was mainly in the training phase, such as teaching ChatGPT to answer questions, which required numerous GPUs (graphics processing units) to process complex data repeatedly. However, AI is now moving into the application phase, where it performs tasks like writing reports, booking flights, and handling emails, requiring real-time interaction with humans and making small decisions. CPUs, being more versatile, are essential for these tasks, especially as the number of such tasks grows exponentially in the agent-based era.
For example, when using an AI assistant to book a flight, it needs to search for flights, compare prices, ask about preferences, and confirm payments—all of which require rapid CPU responses. Many companies are developing such applications, leading to a sudden surge in CPU demand that the supply chain has not been able to meet, resulting in shortages and price increases.
2. Why will Intel outperform NVIDIA by 2026? —— The focus of AI is shifting, with CPUs becoming the key
NVIDIA has thrived in the AI training phase due to its GPU technology. However, by 2026, the focus of the AI industry will shift from training to application, as trained models need to be deployed in various scenarios (such as business offices, smart homes, and autonomous driving). The demand for CPUs will far exceed that for GPUs at this time.
Intel is a established leader in the CPU market with extensive technical expertise and large production capacity. It is also developing CPUs specifically designed for AI applications, which means its chips are likely to be more sought after than NVIDIA's by then. In the application phase, what is needed are CPUs capable of performing tasks, not just GPUs used for training.
3. Intel's stock price soared by 8 times —— The market is betting on future growth
Stock prices reflect investors' confidence in a company's future prospects. Intel's stock has been relatively stable compared to NVIDIA due to its GPU limitations. However, with the rise of AI applications, it has become clear that Intel's CPUs are crucial for the implementation of AI. If 8 out of 10 future AI applications rely on CPUs, Intel's performance is expected to soar. Investors have seen this trend and have bought its stocks, driving up the price—this eightfold increase essentially represents an early recognition of Intel's potential future earnings.
4. What changes will ordinary people experience as AI moves from training to application? —— AI assistants will become more intelligent, but product prices may rise
For consumers, the most direct change will be in the improved functionality of AI assistants. While current AI systems can only answer questions, future agents will be able to complete entire tasks, all powered by CPUs.
On the other hand, CPU shortages and price increases may affect the cost of electronic products, such as computers and smartphones, as well as smart home devices (like speakers and cameras with AI capabilities). However, in the long run, the widespread adoption of AI will make our lives and work more efficient, making these price increases worthwhile.
5. The underlying logic behind the surge in CPU demand during the agent era —— AI needs CPUs to function autonomously
Agents are self-thinking, self-acting AI assistants that require real-time processing for tasks like scheduling work, coordinating with colleagues, and adjusting schedules. GPUs excel at large-scale parallel computing, while CPUs are better suited for flexible sequential processing required by agents. As the number of agents increases, so does the demand for CPUs.
In summary, this news highlights that the "second half" of the AI industry (the application phase) is upon us, with CPUs replacing GPUs as the primary technology. Intel is poised to benefit significantly from this shift, leading to a surge in its stock price and demand. For consumers, this means more practical AI applications, but potentially higher prices for electronic products.