Semiconductors, or chips, each containing countless microscopic circuits, have been around for years as the building blocks that make advanced technology possible. Artificial intelligence (AI) hasn’t necessarily brought anything new to the table in that regard.

It’s artificial intelligence‘s staggering thirst for computing power, and thus, semiconductors, that has created perhaps the most significant investment opportunity the technology sector has seen since the early years of the internet.

Ironically, two of the best AI stocks are of companies that were already known to Wall Street. These two stocks have already turned $10,000 investments into millions over their lifetime, and could continue to thrive as AI chip demand continues to soar.

Here is what you need to know.

Image source: Getty Images.

Nvidia continues to be the gold standard in AI model training

Most technology investors are familiar with Nvidia‘s (NVDA 0.72%) immense success and gains, particularly since the emergence of generative AI in late 2022 with the arrival of ChatGPT. The company’s expertise in GPU chips, which excel at efficiently performing heavy computing workloads, helped Nvidia grab control of the market for data center chips used for training AI models.

More than two years after the AI movement took off, Nvidia still dominates. Research estimates peg Nvidia’s share of AI data center ships for generative AI models at a breathtaking 92%. Of course, the pressure is on Nvidia to hold its share, but it has succeeded thus far in doing so. The company’s CUDA programming platform optimizes the performance of its AI GPU chips, which has been a competitive moat that other companies haven’t breached.

Nvidia succeeded its Hopper architecture with Blackwell, and future chips will continue to deliver more power while consuming less energy. Over time, Nvidia may need to develop new revenue streams beyond data centers. Nvidia will have opportunities as AI eventually extends outside centralized data centers to the edge, bringing computing power to localized AI uses, such as autonomous vehicles and robotics.

For now, AI data center investments continue to roar. Analysts estimate Nvidia will grow revenue by nearly 54% to $199.7 billion this fiscal year, reaching $250.8 billion the following year. Meanwhile, Wall Street anticipates that Nvidia will grow its earnings by an average of almost 29% annually over the long term. While the stock’s forward price-to-earnings (P/E) ratio of 33 is higher than the broader market, it’s a perfectly reasonable price for such a critical AI player with such excellent growth prospects.

Broadcom is growing on strong AI networking momentum, with emerging opportunities in custom chips

Broadcom (AVGO -1.07%) is quickly becoming a formidable force in the AI field. The company built its business on chips for networking and communications, and that’s bearing fruit in the AI space. AI data centers utilize many thousands of chips that operate as clusters. For these clusters to function, all those chips must connect and work together, which requires special networking chips that Broadcom provides.

AI networking chips accounted for approximately 40% of its AI revenue in Q2, representing a 70% year-over-year growth. Similar to Nvidia, Broadcom will likely sell more advanced chips over time that enable companies to build larger clusters of AI chips. Broadcom is diversified well beyond AI, so while the total business probably won’t grow like Nvidia has, it’s also not as dependent on AI, either.

Broadcom’s networking expertise is also unlocking additional AI opportunities. The company has established partnerships with several AI hyperscalers (the companies building all these data centers) on product roadmaps for custom AI chips for model training and inference. Inference chips, which enable models to operate effectively in real-world applications, will become increasingly important as AI continues to mature.

Analysts are also optimistic about Broadcom’s future, with expectations for revenue to grow by 21.7% to $62.7 billion this fiscal year, followed by $74.6 billion the following year. That top-line growth could translate to earnings growth averaging 23% annually over the long term. The stock is a bit more expensive, with a forward P/E ratio of 37, so investors who decide to invest may want to save some cash for when inevitable market volatility creates better buying opportunities.