The VanEck Semiconductor ETF (SMH), a fund tracking the 25 largest U.S.-listed chip companies, has surged 65% year-to-date in 2026, riding a wave of unprecedented demand for AI chips from data center operators worldwide.
As of mid-July, SMH reached approximately $612 per share, up from around $360 at the start of the year, according to multiple financial data sources. The ETF’s explosive performance reflects a broader semiconductor industry renaissance driven by massive capital spending on artificial intelligence infrastructure.
The surge is rooted in concrete spending commitments from major tech firms. According to Goldman Sachs, AI capital expenditure is projected to reach $765 billion in 2026, with much of that directed toward chips and data center infrastructure. Tech companies including Amazon, Microsoft, Alphabet, and Meta are collectively spending over $600 billion on AI infrastructure this year alone, with chips representing a substantial portion of those costs.
SMH’s top holdings—NVIDIA at roughly 19% of the fund, Taiwan Semiconductor Manufacturing Company at approximately 9%, and Broadcom at around 6%—are direct beneficiaries of this spending surge. These companies supply the processors and networking gear essential for training and running large language models.
The semiconductor industry’s revenue is on track to reach $1.3 trillion in 2026, up from $791.7 billion in 2025, according to industry forecasts. AI chip sales alone are projected to approach $500 billion, representing roughly half of total semiconductor revenue. This expansion reflects not only data center demand but also growing AI workloads across consumer devices and enterprise infrastructure.
The 2026 rally builds on momentum from 2025, when SMH returned roughly 49%, and 2024, when it gained 39%. However, the year-to-date surge is notably steeper, signaling accelerating demand as hyperscalers race to secure capacity for AI services.
Data center spending is expected to hit $750 billion in 2026, with AI workloads consuming an estimated 70% of global memory chip production. This concentration of demand on memory and logic chips has created supply pressures across the entire semiconductor supply chain, benefiting equipment makers and foundries alongside chip designers.
Sources
- VanEck — SMH fund holdings and asset allocation
- Yahoo Finance — SMH performance data and year-to-date returns as of July 9, 2026
- Goldman Sachs — AI capital expenditure projections for 2026
- Accuris Tech — AI data center spending and memory chip demand forecasts
- ETF Database — Global semiconductor sales figures for 2025 and 2026 projections
- TheStreet — Global semiconductor sales data for Q1 2026
