About 30 years ago, the Internet began to go mainstream and completely changed the way businesses operated. Although it took some time for companies to fully realize the potential of business-to-business e-commerce and the power of virtual storefronts, the Internet was a real game-changer for American businesses.

Since the mid-1990s, Wall Street and investors have been waiting for the next breakthrough technology, innovation or trend that will give companies the next leap forward. Artificial intelligence (AI) appears to have answered the call.

Image source: Getty Images.

The appeal of AI-driven software and systems is that they can become better at their assigned tasks over time and learn new skills without human intervention. In theory, this makes the technology useful in almost every sector and industry around the globe. This likely explains why analysts at PwC expect AI to contribute $15.7 trillion to the global economy by the turn of the millennium.

While no publicly traded company has benefited more directly from the AI ​​revolution than Nvidia (NVDA 1.00%)the catalyst most responsible for its gains is expected to disappear in 2025.

The rise of AI has increased Nvidia's valuation by more than $3 trillion

As 2022 came to a close, Nvidia was a $360 billion technology stock on the verge of becoming a leader. Today it is a $3.44 trillion company and is arguably the most important of all technology stocks.

Nvidia's fortunes changed due to its hugely popular AI graphics processing units (GPUs). These are effectively the brains of enterprise data centers, enabling split-second decision making. Orders for Nvidia's H100 GPU (commonly known as “Hopper”) have lagged, while CEO Jensen Huang has called demand for Blackwell's successor GPU architecture “crazy.”

The laws of supply and demand clearly state that if demand for a good or service exceeds supply, prices will rise until demand subsides. Nvidia's Hopper chip has been priced between $30,000 and $40,000, representing a 100% to 300% price advantage over competing AI GPUs. The ability to sell its GPUs at a premium boosted the company's adjusted gross margin to over 75% as of the July 28 quarter-end.

Also, don't overlook the role that Nvidia's CUDA software platform has played. CUDA is the toolkit that developers use to build large language models and squeeze as much processing power as possible out of their Nvidia GPUs. CUDA locks Nvidia's customers into its ecosystem of products and services.

The final piece of the puzzle for Nvidia was its ability to win big orders from Wall Street's most influential companies. At the start of this calendar year, it accounted for approximately 40% of Nvidia's net revenue Microsoft, Metaplatforms, AmazonAnd alphabet. There is no better advertising and advocacy than America's leading companies wanting to use your hardware in their AI-accelerated data centers.

Although Nvidia's rise has been textbook, I believe its biggest competitive advantage will be lost next year.

Nvidia's clearest advantage is likely to disappear in 2025

In one respect, there is no danger of Nvidia losing its computing superiority. The company's Hopper and Blackwell GPUs should have no trouble maintaining their computing edge in AI-focused data centers for the foreseeable future.

However, I expect Nvidia to say goodbye to the advantage in the new year, namely the AI ​​GPU shortage.

As previously mentioned, demand for AI-accelerated chips is exceptionally high. Unfortunately, the supply of these chips is limited. Leading global chip manufacturing company Taiwan semiconductor manufacturing (TSM 2.59%) has increased its chip-on-wafer-on-substrate (CoWoS) capacity, a necessity for packaging high-bandwidth storage needed in AI-accelerated data centers. Throughout 2023 and most of 2024, this increase in CoWoS capacity was simply not enough to meet demand. With order volumes overwhelming, Nvidia has supernatural pricing power.

Image source: Getty Images.

However, it looks like 2025 could be the perfect storm that wipes out the AI ​​GPU shortages and pricing power that have worked in Nvidia's favor.

Based on a report by Taiwanese media and a forecast by analyst Zhan Jiahong of Morgan StanleyTaiwan Semiconductor appears to be ahead of schedule in its efforts to increase CoWoS packaging capacity to 80,000 wafers per month. Instead of achieving this feat in 2026 as planned, it could happen a full year earlier than planned. This means more powerful GPUs are available across the board.

It's also worth remembering that Nvidia is no longer the only rodeo in town. Advanced micro devices is increasing production of its MI300X GPUs and recently unveiled its next-generation MI325X chip, with production scheduled to begin later this year. As external competition increases, the shortage of AI GPUs will decrease.

It's not just external competitors that can put pressure on Nvidia's pricing power. The four “Magnificent Seven” companies, which account for around 40% of net sales, are all developing AI GPUs for use in their respective data centers. While these internally developed chips can't match the processing power of Nvidia's hardware, they are significantly cheaper and more accessible. In other words, there's a value proposition that should lead Microsoft, Meta, Amazon and Alphabet to choose their own AI GPUs over Nvidia's in the coming quarters and years.

This perfect storm will steadily increase the availability of AI GPUs in 2025, reducing the stratospheric pricing power that Nvidia has used to increase its gross margin. If this competitive advantage disappears for Nvidia, it could be nearly impossible for the company's shares to maintain their near-parabolic rise since the start of 2023.