Hyperscalers to Spend Over $500 Billion on AI in 2026: Here Are the 3 Stocks to Buy

According to a recent report published by Goldman Sachs, artificial intelligence (AI) hyperscalers are forecast to spend more than $500 billion on infrastructure this year. As capital expenditure (capex) budgets become larger, developers like Microsoft, Alphabet, Amazon, and Meta Platforms share a common idea: Accelerate data center build-outs.

Let’s dig into three companies poised to benefit from these explosive AI-related infrastructure tailwinds both during 2026 and beyond.

Image source: Nvidia.

1. The jack of all trades: Nvidia

The most obvious beneficiary of accelerating infrastructure spend is Nvidia (NVDA 0.72%). Nvidia kicked off the AI revolution three years ago after big tech realized the company’s graphics processing units (GPUs) could be used to build numerous generative AI applications.

Data by YCharts.

On the surface, relentless demand for Nvidia’s GPUs has fueled an unprecedented rise in the company’s top line. But Nvidia’s profitability profile should be what smart investors are focusing on.

As the company’s operating cash flow continues to grow, Nvidia has positioned itself to double down on its innovation road map — releasing new GPU architectures about every 18 months. While the company’s current Blackwell series is considered the gold standard, Nvidia is already building a massive backlog — reportedly in the hundreds of billions of dollars — as hyperscalers rush to secure the company’s newest chips, dubbed Rubin.

Today’s Change

(-0.72%) $-1.39

Current Price

$191.12

Key Data Points

Market Cap

$4.6T

Day’s Range

$189.47 – $194.49

52wk Range

$86.62 – $212.19

Volume

5.6M

Avg Vol

182M

Gross Margin

70.05%

Dividend Yield

0.02%

This is all to say that as investments in training and inference continue to skyrocket, demand for Nvidia’s general-purpose chipsets doesn’t appear to be diminishing in the slightest.

2. The networking specialist: Broadcom

There is more to building AI data centers than just outfitting these facilities with rows of GPU clusters. What makes AI chips such a hot commodity is their ability to process extremely sophisticated algorithms at high speeds. However, there are underlying challenges to these technological dynamics.

Broadcom (AVGO +0.15%) supplies the nuts and bolts that keep GPU clusters running efficiently around the clock. What I mean by that is developers need to allocate meaningful budget spend to less glorious products such as networking switches and interconnects.

Today’s Change

(0.15%) $0.49

Current Price

$331.22

Key Data Points

Market Cap

$1.6T

Day’s Range

$328.34 – $338.11

52wk Range

$138.10 – $414.61

Volume

1M

Avg Vol

30M

Gross Margin

64.71%

Dividend Yield

0.73%

Another way Broadcom benefits from rising infrastructure spend is through custom silicon. Major companies such as Apple, ByteDance, Alphabet, and Meta reportedly collaborate with Broadcom to design custom application-specific integrated circuits (ASICs).

This is a fancy way of saying that the hyperscalers are looking to complement their existing GPU stacks with their own architectures as a way to lower costs and migrate away from an overreliance on a single provider.

While Broadcom doesn’t find its name in the headlines as much as Nvidia or Advanced Micro Devices, the company’s diversified product suite is just as important to the overall AI narrative. As big tech buys more GPUs, builds more data centers, and introduces next-generation products, Broadcom is in a lucrative position to win incremental business alongside its peers thanks to the rapid expansion of infrastructure capacity.

3. The pick-and-shovel giant: Taiwan Semiconductor Manufacturing

I can’t think of a better company that stands to benefit from the chip demand spike than Taiwan Semiconductor Manufacturing (TSM 2.66%). If Nvidia and Broadcom contain the chips driving the AI revolution, TSMC holds the keys to get the car started.

TSMC is the largest chip manufacturer in the world as measured by revenue — holding an estimated 70% market share. Nvidia, AMD, Broadcom, Micron Technology, and some of the hyperscalers outsource their chip production to TSMC’s foundry.

Taiwan Semiconductor Manufacturing

Today’s Change

(-2.66%) $-9.03

Current Price

$330.52

Key Data Points

Market Cap

$1.7T

Day’s Range

$329.11 – $339.69

52wk Range

$134.25 – $351.33

Volume

510K

Avg Vol

13M

Gross Margin

59.02%

Dividend Yield

0.93%

This role makes TSMC a pick-and-shovel provider for the AI infrastructure boom. In other words, it doesn’t really matter which chip architectures are in high demand. So long as enterprises are spending on AI chips, odds are that TSMC is the company making them. From a macro point of view, rising capital expenditure budgets could be seen as a proxy for TSMC’s trajectory, given the company’s leading role in bringing AI hardware to life.

Per the company’s fourth-quarter earnings report, TSMC’s management sees AI as a generational growth trend — one that should produce robust revenue and profit margin expansion throughout the rest of the decade.

Against this backdrop, TSMC might be the biggest beneficiary of all when it comes to the secular themes fueling infrastructure — making it potentially the safest AI chip stock in the market right now for long-term investors.