Alphabet Stock Investors Just Got Great News From a Wall Street Analyst. It's Bad News for Nvidia.

Alphabet (NASDAQ: GOOGL) (NASDAQ: GOOG) and Nvidia (NASDAQ: NVDA) are two of the most valuable stocks in the world, with market capitalizations of $4.6 trillion and $5.1 trillion, respectively. Both companies have deep roots in artificial intelligence (AI), though there has historically been little overlap between the two.

Alphabet dominates the digital advertising market, due in part to proprietary AI models that rank content and target advertising across YouTube and Google Search. Meanwhile, Nvidia dominates the AI infrastructure market because it brings together superior hardware and a robust software ecosystem.

However, Alphabet has recently become a serious threat to Nvidia. The company is expanding access to its custom AI accelerator chips called Tensor Processing Units (TPUs), which serve as an alternative to Nvidia graphics processing units (GPUs).

Here's what investors should know.

Image source: The Motley Fool.

Alphabet is expanding its addressable market by delivering TPUs directly to customers

Alphabet introduced its first generation of TPUs in 2015. TPUs are custom chips specifically developed to accelerate AI workloads, while Nvidia GPUs are built as general-purpose accelerators. Despite being less flexible (i.e., they run fewer algorithms), Alphabet TPUs are often more cost-efficient for running machine learning applications.

"If companies want to diversify away from Nvidia, TPUs are a good way to do it, and there's a lot of reason to be optimistic," according to Gil Luria, head of technology research at DA Davidson. He estimates TPUs could eventually account for 20% of AI infrastructure sales, making the custom chips a $900 billion business for Alphabet.

Alphabet is already capitalizing on that opportunity. Cloud revenue has accelerated in four straight quarters, driven by strong demand for TPUs and Gemini models. Alphabet has also landed major TPU deals with Anthropic and Meta Platforms, and Morgan Stanley analysts expect TPU sales to increase at 60% annually through 2028.

Importantly, Alphabet has historically rented TPUs to customers through Google Cloud, but the company is changing tack to expand its addressable market. CEO Sundar Pichai says, "As TPU demand grows from AI labs, capital markets firms, and high-performance computing applications, we will begin to deliver TPUs to a select group of customers in their own data centers."

Nvidia still has a durable competitive advantage in its CUDA software ecosystem

That is bad news for Nvidia, but investors need to keep the situation in perspective. Nvidia still has a deep competitive moat in its CUDA programming model, the platform developers use to write code for GPUs. The company has built hundreds of code libraries and application frameworks atop CUDA, making its software ecosystem far more robust than Alphabet's AI software stack.

Additionally, Nvidia introduced CUDA nearly two decades ago. In that time, millions of developers have been trained on (and become comfortable with) the platform, and switching programming models is complex and expensive. So, with a broader software ecosystem and a larger developer base, Nvidia is well positioned to retain its dominance in AI infrastructure even if Alphabet gains share.

Nvidia GPUs currently account for over 80% of AI accelerator sales, and Morgan Stanley thinks the company will still have 70% market share by 2030. Meanwhile, Morgan Stanley estimates custom chips like TPUs will account for about 24% of AI accelerator sales.

More broadly, CEO Jensen Huang believes AI infrastructure spending -- which includes central processing unit and networking sales, two categories where Nvidia has booming businesses -- could increase at 32% annually to reach $4 trillion per year by 2030. Nvidia should have no trouble matching that growth trajectory, given its status as the industry standard in AI infrastructure.

Wall Street says Nvidia stock is a better buy than Alphabet stock

Wall Street's consensus estimate says Nvidia's earnings will increase at 43% annually over the next three years. That makes the current valuation of 32 times earnings look cheap. Indeed, most analysts think the stock is deeply undervalued. The median price target of $297 per share implies 40% upside from the current share price of $212.

Meanwhile, Alphabet has a median price target of $430 per share, which implies 13% upside from the current share price of $380. Analysts see less upside here because they anticipate slower earnings growth of 16% annually over the next three years, which makes the current valuation of 29 times earnings look much less attractive.

Consider this: Price-to-earnings-to-growth (PEG) ratios are calculated by dividing a stock's P/E by projected earnings growth. Alphabet has a somewhat expensive PEG ratio of 1.8, while Nvidia has a very inexpensive PEG ratio of 0.7. That makes Nvidia look particularly attractive at current prices.

Trevor Jennewine has positions in Nvidia. The Motley Fool has positions in and recommends Alphabet, Meta Platforms, and Nvidia. The Motley Fool has a disclosure policy.