Several years ago, semiconductor ETFs were often viewed as a Nvidia (NVDA) story. With Nvidia driving much of the group’s growth and performance, some investors questioned the need for broader exposure. But 2026 has shown why diversification can make a difference. While Nvidia remains a significant AI leader, other semiconductor stocks have outperformed this year as the AI trade broadens into memory, networking, custom chips, foundries, and equipment. For investors, it is a reminder to look beyond the headlines and understand what is under the hood of a semiconductor ETF.
Semiconductors Still Powering Through
Semiconductors have moved from a cyclical technology story to one of the most important pieces of the biggest investment stories: artificial intelligence, cloud computing, defense technology, and digital infrastructure. In 2024 and 2025, investors largely viewed semiconductor ETFs through AI demand, especially Nvidia, advanced chips, and data centers. Now, semiconductor industry growth is increasingly being driven by a smaller number of higher-value chips used in AI training, AI inference, advanced networking, high-bandwidth memory, and data center infrastructure. That creates a long-term demand story, but it also raises issues about valuations, competition, supply constraints, and whether AI spending can keep expanding at the same pace.
So far, the industry’s growth has been substantial. According to the Semiconductor Industry Association (SIA), global semiconductor sales reached $791.7 billion in 2025, which is up 25.6% from 2024. SIA also said annual sales are projected to reach roughly $1 trillion worldwide in 2026. Logic and memory were the two largest growth areas in 2025, with logic sales rising 39.9% and memory sales rising 34.8%. The momentum has continued into 2026. SIA reported that global semiconductor sales were $298.5 billion in the first quarter of 2026, up 25% from the fourth quarter of 2025. March sales alone reached $99.5 billion, up 79.2% y/y.
Other Winners Emerge in the Space
As broader semiconductor outlook remains strong, the question is how to play the opportunity. Nvidia remains the dominant AI chip leader, but the trade has started to broaden as investors look for the next layer of beneficiaries across CPUs, networking, memory, foundries, and semiconductor equipment. That is where ETFs can be useful. When a theme is strong but leadership is spread across multiple companies, an ETF can provide diversified exposure to the broader value chain rather than relying on a single stock.
AMD’s latest results highlight this shift. The company reported first-quarter 2026 revenue of $10.3 billion, up 38% year over year, and non-GAAP diluted EPS of $1.37. Its Data Center segment grew 57% year over year to $5.8 billion, driven by demand for EPYC processors and the continued ramp of Instinct GPU shipments. AMD also guided for second-quarter revenue of approximately $11.2 billion, implying around 46% y/y growth at the midpoint. In AMD’s earnings release, management stated that “Data Center (is) now the primary driver of our revenue and earnings growth.” AMD also pointed to growing demand for its high-performance CPUs and accelerators.
Wide Array of Semiconductor Opportunities Exist
There are several semiconductor ETFs to choose from, including leveraged products.
The VanEck Semiconductor ETF (SMH) is the largest ETF of the group with almost $60 billion in assets. It also has the most exposure to Nvidia (a 17% weighting). It tracks an index of the 25 largest and most liquid U.S.-listed companies (including depositary receipts) in the semiconductor industry. Constituents must generate at least 50% of their revenue from semiconductors. VanEck also offers the VanEck Fabless Semiconductor ETF (SMHX), which includes only fabless companies (asset-light companies that focus on chip innovation while outsourcing manufacturing). This means that the ETF excludes some big names like Taiwan Semiconductor (TSM), Texas Instruments (TXN), and Intel Corp (INTC).
The iShares Semiconductor ETF (SOXX) is the oldest ETF of the group and is relatively straightforward. Like SMH, it follows a market-cap weighted index with slightly more (30) constituents. Holdings among the two look very similar besides some of the smaller holdings. One of the main differences, however, is capping. While SMH holds 17% of its weight in Nvidia, SOXX has approximately 7% weight Nvidia with higher allocations to Advanced Micro Devices, Broadcom (AVGO), and Micron (MU).
The SPDR S&P Semiconductor ETF (XSD) follows a modified equal-weight index, which gives it the least exposure to Nvidia among its peers and more exposure across smaller stocks.
Alternative Selections
Unlike the peers that we’ve discussed so far, the First Trust Nasdaq Semiconductor ETF (FTXL) uses a more complex modified factor weighted index which ranks securities based on: 1) trailing 12-month return on assets, 2) trailing 12-month gross income, and 3) momentum. These are weighted based on trailing 12-month cash flow with a max cap of 8%.
The Invesco Semiconductors ETF (PSI) contains 30 U.S. semiconductor companies that were selected based on factors like price momentum, earnings momentum, quality, management action, and value. Invesco also offers the Invesco PHLX Semiconductor ETF (SOXQ), which is another ETF of 30 U.S. semiconductor companies selected by market cap.
Smaller Scale & Leveraged ETFs
Other smaller semiconductor ETFs include the Strive U.S. Semiconductor ETF (SHOC) and the Columbia Semiconductor and Technology ETF (SEMI) (the only active ETF of the group). Global X also released an ETF in October 2025 called the Global X AI Semiconductor & Quantum ETF (CHPX), which takes a global view of semiconductor and quantum computing tech companies. Notably, there are also several leveraged semiconductor ETFs. The largest is the Direxion Daily Semiconductors Bull 3X Shares ETF (SOXL), which aims to produce 300% performance of the NYSE Semiconductor Index (the same index that SOXX tracks). Its short counterpart is the Direxion Daily Semiconductors Bear 3x Shares ETF (SOXS). There are also several single-stock ETFs targeting semiconductor stocks like the GraniteShares 2x Long NVDA Daily ETF (NVDL) and the Graniteshares 2x Long AMD Daily ETF (AMDL).
For more news, information, and analysis visit the Thematic Investing Content Hub.
