Nvidia’s GPUs dominated the AI training market.
But Broadcom’s custom ASICs could conquer the AI inference market.
Nvidia (NASDAQ: NVDA), the world's largest producer of data center GPUs, has been one of the biggest beneficiaries of the AI boom. The world's top AI companies use its GPUs to train their large language models (LLMs), and it locks in its customers with its proprietary software.
From fiscal 2022 to fiscal 2026 (which ended this January), Nvidia's revenue and EPS grew at CAGRs of 68% and 89%, respectively. That's why its stock surged more than 1,000% over the past five years. It's still a great long-term AI play, but another tech giant might generate even more impressive gains over the next few years as the AI market enters a new phase.
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 »
Image source: Getty Images.
Over the past few years, many AI companies have focused on training their algorithms by feeding them massive amounts of data. These training sessions, which can run for weeks to months, use large clusters of Nvidia's GPUs in data centers.
But today, inference -- or the act of using software to access that trained data -- is becoming the next big growth phase of the AI market. As hyperscalers try to reduce the high costs of running their AI models around the clock, they're shifting away from Nvidia's general-purpose GPUs toward specialized application-specific integrated circuits (ASICs).
Instead of being jacks-of-all-trades, ASICs are customized for specific tasks. At scale, these specialized chips can be more cost-efficient than Nvidia's GPUs for handling inference tasks. Therefore, Broadcom (NASDAQ: AVGO) -- which controls about 70% of the ASIC market -- will likely experience a major growth spurt as more AI companies ramp up their investments in inference-driven technologies.
Broadcom already produces custom ASICs for Alphabet's Google, Meta Platforms, OpenAI, and Anthropic. In fiscal 2025 (which ended last November), its AI chip sales soared 65% to $20 billion, accounting for 31% of its top line. It expects that figure to rise to $100 billion in fiscal 2027.
From fiscal 2025 to fiscal 2027, analysts expect Broadcom's revenue and adjusted earnings before interest, taxes, depreciation, and amortization (EBITDA) to both grow at CAGRs of about 53%. Its soaring AI chip sales should offset slower sales of non-AI chips (including wireless, mobile, data center, storage, and industrial markets) and infrastructure software (which it increased its exposure to over the past decade via several major acquisitions).
Broadcom already has an enterprise value of $1.98 trillion, but it trades at just 16 times next year's adjusted EBITDA. That low valuation will attract more investors seeking the next big AI stock -- and Broadcom could be a better play on the inference market than Nvidia.
Before you buy stock in Broadcom, 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 Broadcom 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 $417,305!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,293,148!*
Now, it’s worth noting Stock Advisor’s total average return is 936% — a market-crushing outperformance compared to 209% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
See the 10 stocks »
*Stock Advisor returns as of June 22, 2026.
Leo Sun has positions in Meta Platforms. The Motley Fool has positions in and recommends Alphabet, Broadcom, Meta Platforms, and Nvidia. The Motley Fool has a disclosure policy.