Over the past few years, demand for artificial intelligence (AI) chips has consistently outstripped supply. That imbalance was caused by the soaring popularity of large language models (LLMs), generative AI applications, and specialized AI agents.
According to Precedence Research, the global AI chip market could expand at a CAGR of 27.9% from 2026 to 2035 as that technological shift continues. To profit from that secular trend, investors can consider buying three of the market's most closely followed AI chip stocks: Nvidia (NVDA 1.36%), AMD (AMD 3.20%), and Broadcom (AVGO 1.21%).
Image source: Getty Images.
The differences between Nvidia, AMD, and Broadcom
Nvidia and AMD both produce discrete GPUs, which can process a wide range of parallel tasks. That makes them better suited for handling graphics applications, LLMs, machine learning tasks, and AI applications than CPUs, which are optimized for sequential tasks.
Nvidia and AMD both initially developed most of the GPUs for gaming PCs. Yet over the past decade, both chipmakers have launched high-end data center GPUs for processing AI tasks. Nvidia generally sells pricier, more powerful GPUs than AMD, which sells more affordable ones.
NASDAQ: NVDA
Key Data Points
As of this writing, Nvidia's H100 data center GPUs cost about $25,000 each, while AMD's comparable MI300X chips cost roughly $15,000 each. Many tech giants use a mix of Nvidia and AMD GPUs, but Nvidia's chips are still typically used for higher-end applications.
Nvidia also locks in its customers with CUDA (Compute Unified Device Architecture), a proprietary programming platform optimized for its own chips. Those applications usually need to be rewritten or modified to work on AMD's GPUs. That's why Nvidia still controls more than 90% of the discrete GPU market, even though it sells the most expensive chips.
NASDAQ: AMD
Key Data Points
Broadcom doesn't develop data center GPUs. Instead, it develops custom AI accelerator chips for hyperscalers like Alphabet's Google and Meta Platforms. It precisely customizes these application-specific integrated circuits (ASICs) for its clients' workloads, making them more power-efficient at a massive scale. Broadcom also bundles its chips with its own networking switches, optical equipment, and infrastructure software to help its customers efficiently expand their data centers. However, its custom chips are much less flexible than Nvidia's and AMD's discrete GPUs.
NASDAQ: AVGO
Key Data Points
Can Nvidia, AMD, and Broadcom all grow in tandem?
Nvidia, AMD, and Broadcom all target many of the same hyperscale customers. However, there could be plenty of room for all three chipmakers to thrive without trampling each other.
According to Fortune Business Insights, the global AI infrastructure market could expand at a CAGR of 29.1% from 2025 to 2032 as more companies use AI to optimize their businesses. As that market grows, Nvidia will continue to serve the general-purpose AI market, AMD will serve cost-conscious customers by bundling together its data center CPUs and GPUs, and Broadcom will help the biggest hyperscale customers support their custom workloads.
So while all three companies are trying to increase their penetration of the data center market, we shouldn't assume Nvidia is the only major maker of picks and shovels for the AI gold rush. Instead, analysts expect all three chipmakers to grow rapidly for the foreseeable future.
From fiscal 2025 (which ended last January) to fiscal 2028, they expect Nvidia's revenue to grow at a CAGR of 47%. From 2024 to 2027, they expect AMD's revenue to increase at a CAGR of 34%. As for Broadcom, Wall Street expects its top line to rise at a CAGR of 38% from fiscal 2025 (which ended last November) to fiscal 2028.
From fiscal 2025 (which ended last January) to fiscal 2028, they expect Nvidia's revenue to grow at a 47% CAGR. From 2024 to 2027, they expect AMD's revenue to increase at a CAGR of 34%. As for Broadcom, Wall Street expects its top line to rise at a CAGR of 38% from fiscal 2025 (which ended last November) to fiscal 2028.
Instead of wondering which of these chipmakers will profit the most from the AI boom, it would be a smart move to simply buy all three. They might experience volatile swings in this choppy market, but they could deliver impressive long-term gains for their patient investors.
