Key Points
-
Alphabet trained its latest AI model, Gemini 3, on its own TPU chips.
-
Amazon has teamed with Anthropic to unleash what will soon be over 1 million of its Trainium2 chips.
-
Microsoft and OpenAI are teaming up to develop their own custom AI chips, too.
The artificial intelligence (AI) market is booming, and could grow from $235 billion last year to $631 billion by 2028. The top AI stocks have already enjoyed tremendous returns since early 2023, and the future looks bright. Remember, the internet hasn’t stopped growing since its initial boom decades ago.
Nvidia has arguably been the most apparent AI winner thus far. Its GPU chips have become the de facto choice for AI hyperscalers training and operating AI models, a Golden Goose that has generated $187 billion in revenue over the past four quarters alone.
Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now, when you join Stock Advisor. See the stocks »
But most of that has come from a relatively small handful of these hypercalers. As spending for all these data center projects adds up, the pressure is building for hyperscalers to cut costs.
These three companies, which are, ironically, Nvidia’s own customers, could kill Nvidia’s Golden Goose. They are threats sitting in plain sight.
Image source: Getty Images.
1. Alphabet
After the recent unveiling of its Gemini 3 AI model, Alphabet (NASDAQ: GOOGL)(NASDAQ: GOOG) might pose the most serious threat to Nvidia’s AI dominance. From most accounts, Gemini 3 is a highly impressive step forward for AI models. But perhaps the most remarkable thing about Gemini 3 is that Alphabet apparently trained it on its own proprietary AI chips rather than Nvidia’s GPUs.
Nvidia’s GPUs use the company’s CUDA programming to efficiently harness the immense computing power of GPUs for AI tasks. However, Alphabet’s Tensor Processing Units (TPUs) are application-specific integrated circuits (ASICs), meaning they are designed and built for one job and one job only. In this case, that’s Alphabet’s AI.
Alphabet’s success in training Gemini 3 with its TPUs shows that Nvidia’s GPUs aren’t irreplaceable. That doesn’t mean other companies will easily do what Alphabet did, but it does send the message that it’s possible. Over time, other hyperscalers could design their own AI ASICs, potentially eating into Nvidia’s business.
2. Amazon
Other hyperscalers have also developed custom AI chips. Amazon (NASDAQ: AMZN) operates AWS, the world’s leading cloud computing services platform. AI has created tailwinds for AWS because it primarily runs in the cloud, as does most modern software. Amazon continues to build data centers to increase its cloud capacity, and those data centers require many chips.
Amazon has developed its own AI chip, Trainium, and has begun to assert it more. The company has a partnership with Anthropic, the developer of the AI application Claude. The two companies recently announced the activation of Project Rainier, a massive AI chip cluster. It’s using nearly half a million Trainium2 chips, and that will scale to over 1 million by the end of this year.
It doesn’t necessarily mean Nvidia is out of the picture, but it loosens Nvidia’s stranglehold on the market. At the very least, Nvidia, which has enjoyed 70% gross margins over the past year, could begin feeling some pressure. Between Amazon’s own chip needs and those of a prominent AI developer such as Anthropic, further leaning into Trainium is another potential headwind for Nvidia’s growth.
3. Microsoft
If you’ve noticed a theme by now, you’d be correct. Some of Nvidia’s biggest customers have begun looking for ways to reduce their dependence on Nvidia’s GPUs. Microsoft (NASDAQ: MSFT) is yet another name on this list for similar reasons. Not only is Microsoft’s Azure the world’s second-leading cloud services platform, but it also has a close partnership with ChatGPT developer OpenAI.
OpenAI is the leading AI developer, as ChatGPT is the most popular AI app. However, the company is operating at heavy losses and has recently faced intensifying scrutiny over how it will fund $1.4 trillion in AI infrastructure it has signed agreements for. It seems that the days of writing blank checks for GPUs are nearly over.
Microsoft and OpenAI recently restructured their partnership and will work more closely on custom AI chips. OpenAI has been designing custom chips with Broadcom, and Microsoft CEO Satya Nadella recently suggested that Microsoft will contribute to those efforts so that both companies can benefit. Once again, it’s a shot across Nvidia’s bow, a signal that its lucrative dominance in AI data center chips may soon face serious challenges.
Should you invest $1,000 in Alphabet right now?
Before you buy stock in Alphabet, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Alphabet wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004… if you invested $1,000 at the time of our recommendation, you’d have $580,171!* Or when Nvidia made this list on April 15, 2005… if you invested $1,000 at the time of our recommendation, you’d have $1,084,986!*
Now, it’s worth noting Stock Advisor’s total average return is 1,004% — a market-crushing outperformance compared to 194% for the S&P 500. Don’t miss out on the latest top 10 list, available when you join Stock Advisor.
*Stock Advisor returns as of November 24, 2025
Justin Pope has positions in Alphabet. The Motley Fool has positions in and recommends Alphabet, Amazon, Microsoft, and Nvidia. The Motley Fool recommends Broadcom and recommends the following options: long January 2026 $395 calls on Microsoft and short January 2026 $405 calls on Microsoft. The Motley Fool has a disclosure policy.