The strategy reflects a view that the AI investment opportunity extends beyond the headline technology companies developing AI models. As demand for AI applications grows, the infrastructure required to train and run those models is also expanding, creating opportunities across the semiconductor and data-centre ecosystem.
Within the 22.4% core AI franchise allocation, accelerators and custom silicon account for 8%, analog and power semiconductors for 6.1%, equipment, materials and design for 4.2%, foundry and advanced packaging for 3% and optical and
PCB manufacturing for 1.1%
The fund has exposure to companies such as NVIDIA and Broadcom in accelerators and custom silicon, while ASML, Applied Materials and Cadence feature in equipment, materials and design. TSMC and GlobalFoundries are among the holdings in foundry and advanced packaging.
Compute, networking, memory and power
The remaining 17.9% of assets is spread across AI infrastructure bottlenecks. Compute accounts for 7.2%, networking 5.1%, memory 3.5% and power 2.1%.
The presentation identifies these areas as potential beneficiaries as AI demand increasingly runs into physical infrastructure constraints. Data-centre systems spending is forecast to rise 62.5% in 2026, while the optical transceiver market is expected to grow 65%.
Power is another emerging requirement for AI infrastructure. The fund presentation notes that AI-focused data-centre power use is poised to triple, while data-centre electricity consumption is expected to rise significantly by 2030.
Software and cybersecurity add to AI exposure
The fund is also using a broader technology approach rather than concentrating solely on semiconductor companies. Enterprise software and data account for 18.4% of the portfolio, while cybersecurity and edge account for 10.1%.
The strategy is built around a high-conviction portfolio. The underlying JPMorgan US Technology Fund held 68 stocks as of August 31, 2026, with an active share of 72%. Its top holdings included Palo Alto Networks, NVIDIA, Broadcom, Salesforce, CrowdStrike, Microsoft and Snowflake.
A different approach from Nasdaq-100
The fund's approach also differs from a conventional Nasdaq-100 allocation. The presentation says only around 38% of the portfolio overlaps with the Nasdaq-100 by weight, while the fund has lower exposure to the Magnificent Seven.
For Indian investors, the strategy provides access to technology segments that have limited representation in domestic equity markets, including AI chip designers, semiconductor equipment makers and hyperscale cloud platforms.