The Invesco Dorsey Wright Technology Momentum ETF (NASDAQ:PTF) has gained 58% year-to-date in 2026, a direct result of its rules-based strategy capturing the historic capital spending boom in artificial intelligence infrastructure.
"Investors are looking for ways to play the AI buildout beyond a handful of mega-cap chip designers," said a senior analyst at a technology-focused hedge fund. "Thematic ETFs that track momentum are systematically identifying the secondary and tertiary beneficiaries of this spending cycle, from networking to materials."
The fund’s performance is anchored in the massive expenditures by Big Tech. Amazon, Microsoft, Alphabet, and Meta are on track to spend a combined $725 billion on capital projects this year, with a large portion allocated to building out AI-ready data centers. This spending spree has ignited demand across the entire semiconductor supply chain, far beyond just the GPUs that power AI models.
This infrastructure buildout represents a multi-trillion-dollar opportunity that is becoming mandatory for companies to remain competitive. While Nvidia remains the primary beneficiary with its dominant market share in AI accelerators, the sheer scale of investment is lifting companies that provide the essential "picks and shovels" for data centers.
Beyond GPUs: The ‘Picks and Shovels’ of the AI Boom
The AI boom is creating enormous demand for components that enable the rapid, efficient movement of data. AI servers require advanced compound semiconductors for fiber-optic connections that minimize heat and power loss. This has thrust companies like AXT (NASDAQ:AXTI), a manufacturer of indium phosphide and gallium arsenide substrates, into the spotlight.
AXT’s substrates are critical for the optical networking hardware that connects racks of GPUs, without which data centers cannot function efficiently. The company’s revenue rose 38.6% year-over-year in its last quarter to $26.9 million, and its backlog hit a record $100 million, signaling sustained future demand. This highlights a broader trend where niche suppliers are becoming central to the AI investment thesis. Similarly, Broadcom (NASDAQ:AVGO) is positioned to capture over $100 billion in revenue from custom AI chip designs by 2027 as hyperscalers seek to create more cost-efficient, specialized processors.
Nvidia’s Next Frontier: AI in Space
While the terrestrial data center buildout is the primary driver of current returns, leading players are already expanding the horizon. Nvidia (NASDAQ:NVDA), the leader in AI chips, is positioning itself to dominate AI applications in space, a market projected to grow to $110.2 billion by 2035 from just $6.2 billion in 2025.
The company has already launched space-hardened computing platforms, such as the Space-1 Vera Rubin Module, to enable in-orbit data processing and autonomous space operations. CEO Jensen Huang has noted that while space-based data centers are not yet economical, the potential for solar-powered, AI-driven applications is a significant long-term growth vector. Companies like Planet Labs (NYSE:PL) are already using Nvidia’s technology for satellite imagery analysis, underscoring the tangible, near-term applications of AI beyond Earth.
This article is for informational purposes only and does not constitute investment advice.
