Artificial intelligence (AI) might be the most revolutionary technology in a generation. Depending on which Wall Street forecast you rely on, AI could add $7 trillion (Goldman Sachs), $15.7 trillion (PwC), or even $200 trillion (Ark Investment Management) to the global economy over the coming decade.
Before all of that value can be realized, Nvidia (NVDA -2.09%) CEO Jensen Huang says the technology industry needs to invest at least $1 trillion in data center infrastructure over the next five years to meet demand from AI developers. That includes chips, networking equipment, and other components, so it presents an incredible opportunity for the entire semiconductor sector.
There's one exchange-traded fund (ETF) with an incredible track record of helping investors profit from hardware booms like this. The iShares Semiconductor ETF (SOXX -0.85%) delivered a return of 794% over the last 10 years alone. Here's how it could turn an investment of $250,000 into $1 million over the next 10 years.
Nvidia is the top holding in this ETF
The iShares Semiconductor ETF aims to give investors exposure to companies that design, manufacture, and distribute chips -- and especially those that stand to benefit from powerful trends like AI. An exchange-traded fund (ETF) can hold hundreds or even thousands of stocks, but this one only holds 30. That means it's highly concentrated, which is a risk investors should keep in mind. It can rise and fall based on the performance of a mere handful of stocks because its top five positions account for 39% of the total value of its portfolio:
Stock |
iShares ETF Portfolio Weighting |
---|---|
1. Nvidia |
10.48% |
2. Broadcom |
8.59% |
3. Advanced Micro Devices |
7.90% |
4. Texas Instruments |
6.16% |
5. Qualcomm |
5.91% |
Nvidia's graphics processing units (GPUs) for the data center are the most sought-after pieces of hardware for AI developers. The company generated a record $30.8 billion in data center revenue during its recent fiscal 2025 third quarter (ended Oct. 27), which represented a whopping 112% growth from the year-ago period. Now, it's ramping up shipments of its new Blackwell GPUs, which offer a substantial leap in performance, and Jensen Huang says demand is "staggering."
Broadcom manufactures AI accelerators (another type of chip) for hyperscale tech giants and also supplies data center components, like ethernet switches, which regulate how quickly data travels between chips and devices. The company is experiencing a surge in demand. During its fiscal 2024 third quarter (ended Aug. 4), sales of its AI accelerators grew by 3.5x, compared to the year-ago period, and sales of its Tomahawk 5 and Jericho3-AI switches soared fourfold.
Then there's Advanced Micro Devices (AMD), which has become one of Nvidia's biggest threats in the data center. The company's latest GPUs have performance and cost advantages over some of Nvidia's hardware, and it's launching a new lineup next year to compete directly with Blackwell. Outside of the data center, AMD is also a leading supplier of AI chips for personal computers, which could be a major long-term growth driver.
Beyond its top five positions, the iShares ETF holds a number of other important chip stocks. They include Micron Technology, a top supplier of memory and storage chips, which are becoming critical to AI development, and Taiwan Semiconductor Manufacturing, which fabricates many of the chips designed by Nvidia and AMD.
Turning $250,000 into $1 million
The iShares Semiconductor ETF has delivered a total return of 1,170% since its inception in 2001, which translates to a compound annual return of 11.6%. That's much better than the average annual gain of 8.2% generated by the S&P 500 over the same period.
As I highlighted earlier, the bulk of that gain has come over the last 10 years, when the ETF's compound annual return accelerated to 24.5%. Before AI came along, tech giants were building data centers to provide cloud computing services to businesses all over the world. Plus, the widespread adoption of smartphones, tablets, and notebook computers also boosted demand for chips and components.
The below table highlights how long it could take for the iShares ETF to turn an investment of $250,000 into $1 million, based on three different performance scenarios:
Starting Balance |
Compound Annual Return |
Time to Reach $1 Million |
---|---|---|
$250,000 |
11.6% |
13 Years |
$250,000 |
18.1% (midpoint) |
9 Years |
$250,000 |
24.5% |
7 Years |
If the iShares ETF maintains its average annual return of 24.5% from the last 10 years, it will easily deliver a fourfold gain for investors within the next decade. However, even if it reverts back to its long-term average return of 11.6% per year, investors could still see their $250,000 grow into $1 million over 13 years.
Maintaining a return of 20% or more per year won't be easy because the law of large numbers eventually catches up. For example, Nvidia will be a $21 trillion company in 10 years if its stock increases by 20% annually. Considering Huang himself only forecasts $1 trillion in AI data center spending over the next five years, it's unclear where that additional value would come from.
With that said, the AI revolution is still in its infancy. Companies like Nvidia and AMD are also investing heavily in their software businesses to complement their powerful chips, which charts a new pathway for them to create value. They might even capture some of the multitrillion-dollar AI forecasts from Wall Street that I mentioned at the beginning of this article.
I think the best innovation in the AI semiconductor industry is still ahead of it, so the iShares ETF could be a great long-term buy for investors. However, it's important to own it as part of a balanced portfolio of other funds or individual stocks, just in case AI doesn't live up to the hype.