The artificial intelligence (AI) market is undergoing significant changes. Centered around NVIDIA in the United States, which once drove the market rally, many stock prices have declined. On the other hand, the share price of Alphabet, the parent company of Google, has risen notably. This is because Google’s generative AI foundational model released in November has received high praise. The market has sensed the possibility that the dominant position of ChatGPT, led by OpenAI in the US, may be shaken.

Alphabet’s stock rose by 14% in November. Its total market capitalization approached $4 trillion, surpassing Microsoft for the first time in six years and ranking third in the world. The contrast with NVIDIA, whose stock price fell by 13%, is quite striking.

The significant change in stock market evaluations is attributed to the expansion of Google’s new generative AI models and the application of its self-developed semiconductors.

On November 18, Google unveiled its latest large language model (LLM) for generative AI, “Gemini 3”. One of the new models, “Gemini 3 Pro”, topped the main indicators for comparing LLM performance, becoming a strong competitor to ChatGPT.

The promotion of Google’s self-designed semiconductor, the “TPU (Tensor Processing Unit)”, which is used in its own AI development, has also contributed. On October 23, Google announced that it would provide products to Anthropic, a US startup company involved in AI development. In late November, there were reports that Meta was considering applying it in its data centers.

It is claimed that compared with NVIDIA’s GPUs, which almost monopolize the market, TPUs have lower costs, strengthening expectations of market share gains. Morgan Stanley in the US estimates that if TPU sales increase by 500,000 units, Alphabet’s earnings per share (EPS) for 2027 will be pushed up by about 3%.

The improved evaluation of Google has also benefited companies in the “Google camp” believed to be closely related to the company. Broadcom, which collaborates with Google on TPU design, rose by 9% in November, and Taiwan’s MediaTek also increased by 6.5%.

From the perspective of Japanese stocks, Toppan Holdings hit a record high in November. The investment from Broadcom in its new semiconductor packaging substrate factory in Singapore, set to start production by the end of 2026, is seen as a positive factor.

In stark contrast is the stock performance of companies in the “OpenAI camp” centered around NVIDIA, which previously drove the AI market rally. Microsoft, an investor in OpenAI, fell by 5%, and SoftBank Group (SBG) dropped by 38%.

The stock market fluctuates based on the competitive landscape among AI companies. In early 2025, news emerged that China’s DeepSeek had developed a high-performance generative AI model at a low cost. Fears that the advantages of US AI companies might be shaken caused a temporary market plunge.

After summer, whenever there was news of strengthened cooperation, such as NVIDIA investing in OpenAI, the stock prices of related companies like Oracle and SoftBank Group rose. In October, NVIDIA’s total market capitalization reached 5trillion,andMicrosoft′salsohit4 trillion.

At the same time, concerns about over-investment by tech companies have surfaced. Against a backdrop of growing skepticism towards the existing winners, Alphabet’s rise has led to the current polarization of stock prices.

NVIDIA’s forward price-to-earnings ratio (PER, based on expected earnings in 12 months) is over 25 times, lower than its average over the past 10 years (over 37 times). Hirokazu Kurachi, a fund manager at Japan’s Asset Management One, pointed out, “Options for AI semiconductors are increasing beyond NVIDIA. This indicates that the market is skeptical about the sustainability of high growth rates.”

The competition in AI development is extremely fierce. Fang Jiayan, a senior fund manager at Mitsui Sumitomo DS Asset Management, believes that competition will intensify and there will be repeated changes in the leading players in the market.

If achievements in new areas such as “physical AI”, which autonomously controls robots and machinery, are confirmed, a new leading player to replace Alphabet is likely to emerge.