Summary of Key Points
On June 1st, the chip sector showed divergence before the U.S. market opened: Nvidia's shares rose due to its investment in AI PCs based on the Arm architecture, while those of Intel and Qualcomm declined due to competitive pressures. Behind this was a clash for dominance in AI infrastructure between Nvidia and Intel—Nvidia is expanding from its position as a GPU leader to a full computing platform provider (including CPUs and PC ecosystems), while Intel is firmly holding onto its stronghold in data center CPUs. The shift towards AI models being deployed on endpoints is driving changes in PC architecture, and it's possible that the future Windows PC market will be dominated by a quartet of players: Intel, AMD, Qualcomm, and Nvidia. Arm is set to benefit the most from this transition as the provider of the underlying architecture licenses. The rules of the chip competition have changed, with the ability to integrate CPUs, GPUs, networking, and software ecosystems becoming the key to success.
I. Divergence in Chip Stocks: Some Gain, Some Lose—The Difference Lies in "Track Selection"
Why did Nvidia's shares rise while Intel's fell? Simply put: Nvidia is entering new territories, while Intel and others are defending their established positions.
- Nvidia's gains: It has partnered with Microsoft, Arm, and MediaTek to develop AI PCs that can run AI models locally (e.g., generating images or performing translations without an internet connection), representing a major transformation for the PC industry. Investors are optimistic about this new strategy.
- Intel and Qualcomm's declines: These companies were traditionally strong in the PC and mobile chip markets, but now Nvidia is entering with its AI advantages, potentially squeezing their market shares. For example, Qualcomm was trying to enter the PC market through the Arm architecture, only for Nvidia to also compete for that share, increasing their pressure.
II. Nvidia's Ambition: From "King of GPUs" to "Controller of the Full Computing Platform"
Nvidia has previously made a fortune in data centers with its GPUs (the core of AI computing). However, it aims to be more than just an AI accelerator; it wants to dominate the entire computing ecosystem:
- Data Centers: It has launched the Grace series of CPUs that integrate GPUs and CPUs with networking, along with its CUDA software (widely used by developers for AI programming), making its solutions highly attractive to customers.
- PC Market: By collaborating with Arm on AI PCs, Nvidia is bringing data center technologies to endpoints—using Arm-based CPUs paired with its own GPUs and Microsoft's Windows operating system to directly compete with Intel in the core PC market.
III. Intel's Counterattack: Defending Its "Lifeline" in Data Centers
Intel's main business lies in CPUs, especially those for data centers (which used to account for a large portion of its revenue). With Nvidia targeting the CPU market, Intel is fighting back fiercely:
- It has released the Xeon 6+ processor, the first data center CPU built using its advanced 18A manufacturing process, tailored for AI tasks (such as handling multiple tasks and fast data transfer), which has already been tested and deployed in telecommunications and data centers.
- Why This Is a Critical Battle: If Intel loses the data center CPU market to Nvidia, it will lose its main source of profit, so it must use its best technology to defend this territory.
IV. The Arrival of the AI PC Era: A Transformation for the PC Market
In the past, PCs were primarily used for office work and gaming. With the advent of AI, PCs need to be capable of running AI tasks locally—e.g., generating content or editing images in real-time without relying on the cloud. This change is reshaping the PC market landscape:
- Quartet Competition: The market was once dominated by Intel and AMD, but now Arm's versatile architecture has entered the picture, along with Nvidia's AI capabilities, creating a four-way competition among Intel, AMD, Qualcomm, and Nvidia.
- Arm's Win-Win Situation: Whether it's Nvidia or Qualcomm's AI PCs, they both use the Arm architecture, allowing Arm to earn licensing fees for each chip sold, making it the biggest beneficiary.
V. New Rules of Chip Competition: "Integration Ability" Is More Important Than "Single Hardware"
In the past, chip companies competed based on factors like CPU speed or GPU performance. However, in the AI era, success depends on integrating CPUs, GPUs, networking, and software efficiently. Nvidia's strength lies in its ability to offer a complete solution (hardware and software), which gives it a significant advantage. Intel, on the other hand, may be at a disadvantage if it only focuses on CPUs without a supporting AI ecosystem.
In short, the future of chip companies lies in forming strong ecosystems rather than competing as individual entities. The ultimate winner will be the one that can combine both hardware and software effectively. This news highlights how AI is reshaping the entire chip industry, with all players realigning their strategies. The winner will be the company that can harness both "hardware" and "software" to its advantage.