As of writing, Intel's stock has surged 22% today, bringing its year-to-date gain to 116%.
Several other related stocks have also posted strong gains this year (format: company, today's gain, year-to-date gain): Advanced Micro Devices (AMD): 14% today, 59% YTD ARM Holdings (ARM): 13% today, 106% YTD Other stocks pulled higher today include Taiwan Semiconductor Manufacturing (TSMC), up 6%, and NVIDIA (NVDA), up 5%.
Let's examine the reasons behind this rally.
The primary driver is the CPU business, specifically not personal computer CPUs but Data Center and AI (DCAI) CPUs.
The core conclusion from a simple analysis is that Intel's year-over-year revenue growth is entirely attributable to its data center CPU segment.
A common question arises: aren't modern data centers dominated by GPUs or ASICs (like TPUs)? Why is CPU demand surging now? The reasons are twofold:
First, the explosive growth in computing power demand has led to a surge in data center construction, driving orders for all related components. This includes not only CPUs but also memory, natural gas generators, and copper.
Second, even early AI data centers focused on pre-training required some CPUs for orchestration and control. With the evolution of AI training methods and the expansion of applications—such as post-training (reinforcement learning), inference, and the emergence of AI agents—the need for high-precision CPU management, planning, and verification has skyrocketed. This is particularly true for AI agent applications.
These two factors are multiplicative, resulting in a sharp increase in CPU demand. Intel's recent earnings report served as powerful confirmation: It demonstrated that data center infrastructure build-out will continue to accelerate, which is why TSMC and NVIDIA's stocks also rose. It proved that CPU demand is indeed rising, which is why AMD and ARM—companies focused on data center CPUs—saw the most significant gains.
Looking ahead, the demand for computing power will continue to grow dramatically due to several factors:
From a competitive standpoint: Major AI model companies are accelerating the launch of leading models to capture a market that is uncertain in size but guaranteed to be massive.
From the perspective of AI profitability: Anthropic's Annual Recurring Revenue (ARR) has skyrocketed from $9 billion to over $30 billion in just three months, an unprecedented increase. OpenAI's revenue growth has also been exceptionally strong.
From the utility perspective of AI: Applications are expanding into diverse areas.
From current market demand: Many users face rate limits from providers like Anthropic. Anthropic's new model, Mythos, lacks the computing capacity to serve general customers, leading to a strategy of temporarily offering it at a premium to a select group of wealthy companies. The rental price for H100 GPUs, launched four years ago, has increased by over 40% in the past six months.
From the perspective of future demand growth: More powerful AI will unlock more applications and use cases.
Demand far outstrips supply, leading to higher product prices, increased total industry profits, and attempts to boost supply. However, given numerous bottlenecks in computing power production capacity (a key reason industry leaders are exploring concepts like space-based data centers), growth in hardware supply will struggle to keep pace with software-driven demand.
This environment ensures the continued accelerated expansion of data centers, driving sales volume and profit margins for related components. CPUs are a key part of this, while memory prices have already multiplied several times over recent months.
Ultimately, determining which stocks to consider depends on how the growing profits of the entire AI industry will be distributed.
In summary: demand is growing, profits are growing, and the critical question is how those profits will be allocated.
Comments