Despite the fanfare, the artificial intelligence (AI) revolution has just begun. With the AI market valued at $189 billion in 2023, the United Nations believes it will become a $4.8 trillion market by 2033.
Companies like Nvidia have already taken advantage of this growth, soaring to multitrillion-dollar market caps. But the two AI businesses below trade at just fractions of that value. Over time, however, we could see one of these stocks surpass Nvidia's market cap, leading to huge gains for patient shareholders.
This is what makes Nvidia so special
Right now, most estimates believe that Nvidia commands somewhere between 70% and 95% of the AI graphics processing unit (GPU) market. GPUs, or graphics processing units, are critical components necessary for training and executing AI models, as well as facilitating many other machine learning tasks. Without GPUs, the AI revolution would not be taking off at nearly the same size or scale. And right now, Nvidia dominates AI-specific GPU sales.
What makes Nvidia's GPUs so special? Two things: early investment and vendor lock-in through its developer suite called CUDA.
Way back in 2006, Nvidia's leadership recognized the importance of programmable infrastructure. That is, they understood that developers would want to customize their chips to optimize for certain parameters, allowing them to process data or run calculations faster and more efficiently than a stock GPU. To address this, Nvidia released Compute Unified Device Architecture (CUDA). This unlocked the power of parallel computing, making its chips more attractive than the competition when it came to performance optimization potential.
Today, many Nvidia customers are using Nvidia products due to CUDA. They've customized their setups from a software perspective around Nvidia's hardware offerings, creating what analysts call "vendor lock-in." This lock-in has granted Nvidia an 80% to 95% market share for AI-related GPUs. It'll be hard to compete with this competitive advantage. But eventually, another chipmaker will break through. And the companies below are my top bets when it comes to both risk and potential upside potential.
Two stocks that could eventually beat Nvidia
The road to toppling Nvidia will be a long one. But over the coming years, I suspect either Intel (INTC +4.42%) or Advanced Micro Devices (AMD +0.71%) could break through.
NASDAQ: AMD
Key Data Points
AMD is arguably in the best position to potentially match Nvidia's AI dominance over the next five years. The company's latest GPUs have performed well against Nvidia's Blackwell chips on benchmark tests. Plus, Nvidia is having difficulty manufacturing enough chips to meet demand, leading to multi-month delays on shipments, giving AMD an ability to more rapidly meet rising demand despite arguably inferior products with less vendor lock-in.
NVDA PS Ratio data by YCharts
Right now, Intel is far behind AMD in terms of catching up with Nvidia. But its market cap and valuation more than reflect that reality. Intel is valued at just $80 billion versus a $140 billion valuation for AMD. Meanwhile, Intel shares trade at just 1.5 times sales versus a 5.6 times sales valuation for AMD. Betting on Intel reaching Nvidia's valuation by 2030 is clearly a long shot. But the company is investing heavily to improve its chips' competitiveness, as well as its overall manufacturing capacity. And late last year it received a multibillion-dollar contract from Amazon for AI chips and another multibillion-dollar contract from the U.S. military.
Which company am I betting on today to catch up with Nvidia? I'm going with AMD. Its chip performance and manufacturing capabilities heftily outpace Intel's for AI GPUs. And with 52% of revenue coming from data centers versus just 25% for Intel, AMD is clearly much more leveraged to the AI economy than Intel. Nvidia's CUDA architecture will remain a strong barrier to competition for years to come. But both AMD and Intel have such cheap relative valuations that both are worth a small, speculative investment, even if the odds of overtaking Nvidia by 2030 remain slim.
