1 Super Semiconductor ETF That Could Turn $400 Per Month Into $1 Million, With Nvidia's Help

Nvidia (NVDA -2.09%) pioneered the graphics processing unit (GPU) in 1999 to render computer graphics for gaming and multimedia purposes.

Since GPUs are capable of parallel processing -- meaning they can seamlessly perform multiple tasks at the same time -- they are also ideal for compute-intensive workloads like machine learning and artificial intelligence (AI) development. That led Nvidia to design new GPU architectures for data centers, and the semiconductor industry is now at the heart of the AI revolution.

Nvidia CEO Jensen Huang believes data center operators will spend $1 trillion building GPU-based AI infrastructure over the next five years. That's an incredible financial opportunity, not only for his company, but for the entire semiconductor industry.

The iShares Semiconductor ETF (SOXX -0.85%) holds every leading chip stock, so it can give investors exposure to that trend in a diversified way. In fact, here's how the exchange-traded fund (ETF) could turn $400 per month into $1 million over the long term.

Image source: Getty Images.

Every top chip stock packed into one fund

The iShares Semiconductor ETF invests in U.S. companies that design, manufacture, and distribute chips -- especially those poised to benefit from powerful trends like AI. Although ETFs can hold hundreds or even thousands of different stocks, the iShares Semiconductor ETF holds just 30, so it's highly concentrated toward its singular theme.

Led by Nvidia, its top five holdings represent 37.9% of the entire value of its portfolio.

Data source: iShares. Portfolio weightings are accurate as of Oct. 14, 2024, and are subject to change.

Nvidia was valued at $360 billion at the start of 2023. Less than two years later, it's now the second largest company in the world, with a market capitalization of $3.2 trillion. The chip giant is delivering the revenue and earnings growth to support its incredible rise in value, thanks primarily to sales of its data center GPUs.

In the recent fiscal 2025 second quarter (ended July 28), Nvidia generated $26.3 billion in data center revenue, which was a whopping 154% increase from the year-ago period. The strong results are likely to continue, because the company is about to start shipping a new generation of GPUs based on its Blackwell architecture. Blackwell GPUs promise an incredible leap in performance of up to 30 times compared to Nvidia's flagship H100 GPU, and Huang recently said demand for them is "insane."

Broadcom also plays a key role in AI data centers. It makes AI accelerators (a type of chip) for hyperscale clients, which typically include tech giants like Microsoft and Amazon. It also makes Ethernet switches like the Tomahawk 5 and Jericho3-AI, which regulate how quickly data travels between GPUs and devices.

Advanced Micro Devices has emerged as a direct competitor to Nvidia in the GPU space. It will ship its new MI350X data center chip, which is designed to compete directly with the Blackwell lineup, in the second half of 2025. But AMD also makes neural processors (NPUs) for personal computers, which can handle AI workloads on-device, creating a faster user experience. This could be a big opportunity for the company outside the data center.

Beyond its top five positions, the iShares Semiconductor ETF also holds other top AI chip stocks like Micron Technology, which supplies memory and storage chips designed increasingly for AI workloads, and Taiwan Semiconductor Manufacturing, which fabricates many of the GPUs designed by Nvidia and AMD.

Turning $400 per month into $1 million

The iShares Semiconductor ETF has generated a compound annual return of 11.6% since its inception in 2001. However, its compound annual return has accelerated to 24.5% over the last 10 years, thanks to the rapid adoption of compute-intensive technologies like cloud computing, enterprise software, and AI.

The table below highlights the returns an investor could earn with $400 per month over 10 years, 20 years, and 30 years based on three different annual growth rates.

Calculations by author.

It's unlikely that the iShares Semiconductor ETF will deliver an average annual return of 24.5% over the next 30 years -- or even over the next 10 years, for that matter. The law of large numbers will eventually lead to a deceleration in growth. Nvidia is experiencing that phenomenon right now. Despite growing its data center revenue by 154% in its recent quarter, that was a much slower growth rate than the prior quarter just three months earlier, when its data center revenue jumped by 427%.

However, even if the ETF reverts back to an annual return of 11.6%, that will still be enough to turn $400 per month into $1 million over 30 years. While nothing is guaranteed, that is a more realistic expectation for investors.

Plus, ETFs can be very flexible. The iShares Semiconductor ETF will rebalance over time, so new companies will find their way into its top holdings if they are outperforming their peers, which will support further returns.

AI is likely to be a game changer for the semiconductor industry over the long term. Goldman Sachs believes the technology will add $7 trillion to the global economy in the coming decade. If that's true, it will drive a consistent reinvestment into chips and infrastructure to fuel future growth cycles.

However, there is always a risk that AI will fail to live up to the hype. That's why it's important for investors to buy the iShares Semiconductor ETF only as part of a balanced portfolio.