Google's "secret weapon"—Will TPU support a $900 billion super track?

```

Investors are increasingly convinced that Alphabet’s artificial intelligence chips could become a major source of revenue for its parent company in the future. These custom chips, seen as Google’s “secret weapon”—the Tensor Processing Unit (TPU)—not only drive internal growth for its cloud computing business, but could also create a nearly trillion-dollar new market by being sold to third parties.

Driven by these expectations, Alphabet’s share price surged by 31% in the fourth quarter, making it the tenth best-performing component of the S&P 500. Optimism about the commercial prospects of the TPU is fueled by a series of recent key developments. At the end of October, Alphabet announced it would provide AI startup Anthropic PBC with billions of dollars’ worth of chips, pushing its share price up over 6% in just two days.

A month later, The Information reported that Meta Platforms Inc. was negotiating to invest billions of dollars in TPUs, causing the stock to jump again. Gil Luria, Head of Technology Research at DA Davidson, estimates that if Alphabet seriously pushes the external sale of TPUs, it could capture up to 20% of the AI chip market within a few years, making it a business worth about $900 billion.

Wall Street generally sees the TPU as Alphabet’s core advantage in the fierce AI race. Even though Alphabet has never sold a single chip externally, more efficient in-house chips mean a more powerful, lower-cost cloud service. But once it begins to seriously sell TPUs to external clients, its business landscape could be fundamentally transformed.

Seeking Alternatives Beyond Nvidia

In today’s AI chip market dominated by Nvidia, the TPU offers companies looking to diversify their supply chains a highly attractive alternative.

Gil Luria, Head of Technology Research at DA Davidson, stated:

“If companies want to reduce their reliance on Nvidia, the TPU is a great choice, which gives us every reason to remain optimistic.”

TPUs are “Application-Specific Integrated Circuits” (ASICs), custom-designed for accelerating machine learning workloads. This specialization makes them less versatile than Nvidia’s chips, but also brings significant cost advantages. Mark Iong, an equity portfolio manager at Homestead Advisers, points out that as investors become more cautious about AI-related spending, cost-effectiveness is a real advantage:

“Nvidia chips are more expensive and harder to get, but if you can use ASIC chips, Alphabet is right there, and far ahead of the market.”

The Trillion-Dollar Imagination Space

Analysts have started to quantify the massive commercial value possible from external TPU sales.

Gil Luria estimates that if Alphabet seriously advances external TPU sales, it could capture 20% of the AI chip market within a few years, making it a business worth around $900 billion.

Morgan Stanley analyst Brian Nowak has also noticed “nascent signs of a TPU sales strategy.” Citing predictions from the firm’s Asia semiconductor analyst, he noted that TPU procurement is expected to reach 5 million units in 2027, around 67% higher than previously expected; and 7 million units in 2028, 120% higher than prior expectations. In a December 1 report to clients, Nowak wrote that while the majority of demand might come from Alphabet’s own use and Google Cloud, this also “reveals Alphabet’s potential to sell more TPUs.”

According to Morgan Stanley’s calculations, every 500,000 TPUs sold to third-party data centers could increase Alphabet’s 2027 revenue by about $1.3 billion and contribute $0.40 to its earnings per share. Based on analysts’ projections, Alphabet’s 2027 revenue is expected to be around $447 billion, so an additional $13 billion would represent nearly 3% sales growth. According to Bloomberg data, market consensus on the company’s 2027 revenue forecast has increased by more than 6% in the past three months.

Synergy from an AI Full-Stack Layout

The TPU’s value lies not only in its potential as a stand-alone product, but in its deep integration within Alphabet’s entire AI ecosystem.

Google’s recently released Gemini AI model has been highly praised and optimized to run efficiently on TPUs, further confirming the intrinsic value of the chip. Mark Iong commented, “Alphabet is the only company with a leading position in every layer of AI,” pointing out that from the Gemini model and Google Cloud to TPUs and other areas, “this gives it incredible advantages.”

Nonetheless, it’s currently unclear how committed Alphabet is to large-scale third-party chip sales, but its internal advantages have already paved the way for future business decisions.

Excessive expectations for the TPU business could disappoint investors if not realized, especially given that Alphabet’s share price has already risen sharply.

Currently, Alphabet’s share price is about 27 times expected earnings, the highest since 2021 and well above the decade average. However, even so, its valuation is lower than that of other big tech stocks like Apple, Microsoft, and Broadcom.

Allen Bond, a portfolio manager at Jensen Investment Management, recently used the share price rally as an opportunity to reduce holdings. However, he remains optimistic about the company’s overall position and prospects, believing “the pathway for TPU to become a revenue driver is credible.” He said:

“Alphabet is showing real strength and progress in AI, and while this is increasingly recognized by investors, given the growth expectations, its valuation still appears reasonable. A company with strong momentum in AI trading at a lower valuation than Microsoft and Apple means it remains a core holding.”

Risk Warning and DisclaimerThe market has risks; investment needs caution. This article does not constitute personal investment advice, nor does it take into account the special investment objectives, financial status, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article fit their specific circumstances. Investment based on this is at your own risk. ```