The AI trade has become a race to find the next NVIDIA Corp (NASDAQ:NVDA) , but Jensen Investment Management’s Allen Bond sees another way to play the boom: own the companies that benefit regardless of which AI model wins.
His picks — Broadcom Inc. (NASDAQ:AVGO), KLA Corp (NASDAQ:KLAC) and Amphenol Corp (NYSE:APH)— sit deeper in the infrastructure stack.
"Continued AI adoption requires more compute, power and bandwidth regardless of which models or platforms ultimately win," Bond, the managing director and portfolio manager at Jensen, told Benzinga in an exclusive email interview. He sees the three companies as durable AI infrastructure plays because of their competitive advantages, mission-critical products and established customer relationships.
• Broadcom stock is facing resistance. What’s going on with AVGO shares?
The AI Trade Gets Broader
That model-agnostic angle is becoming increasingly important as hyperscalers develop their own chips, AI companies pursue different architectures and investors debate which model developers will ultimately capture the economics.
Bond says the more durable opportunities may lie with "high-quality infrastructure providers enabling its proliferation." That includes semiconductor equipment, networking and custom silicon, as well as the hardware needed to connect increasingly complex data-center systems.
Three Ways Into AI
- Broadcom fits that thesis through its custom AI accelerators and networking technology. The company expanded its AI infrastructure portfolio across XPUs, Ethernet, optics, SerDes and PCIe solutions for large-scale clusters.
- KLA offers a different exposure. Its process-control equipment helps semiconductor manufacturers identify defects and improve yields as chips become more complex. Jensen has described KLA as a "clear AI investment beneficiary" while emphasizing that the company also has its own competitive advantages and growth prospects.
- Amphenol brings the infrastructure thesis further down the chain. Its connectors and interconnect systems help move power and data through increasingly dense computing environments, including AI data centers. The company has specifically highlighted demand for faster computing, scalable memory architectures and data movement as AI infrastructure expands.
For Bond, that diversification matters. Instead of relying on one model developer, accelerator architecture or AI application, these businesses can benefit from the broader physical buildout required to support the technology.
That is also consistent with Jensen’s broader investment philosophy. The firm says it looks for businesses with demonstrated competitive advantages and strong returns on capital rather than simply chasing momentum around the AI theme.
See More: Top Growth Stocks
Investment Takeaway
The AI infrastructure trade is getting bigger than GPUs. Bond’s thesis is that compute, connectivity and semiconductor manufacturing remain necessary even if today’s AI leaders or architectures change.
For investors looking beyond Nvidia, Broadcom, KLA and Amphenol offer three different ways to participate in that underlying buildout — without having to pick the ultimate winner of the AI model race.
Image Courtesy: Jensen Investment Management
Login to comment