IXSAR Insights · Investment Perspective · 2026-08-13 · 2 min read · By IXSAR Capital

The Physical AI Revolution: Why I Invest at the Frontier

Physical AI is not a theme. It is the convergence of three hardware cycles with the most capable AI models ever built. Here is why I think this is the defining venture cycle of the decade.

When I joined IXSAR as General Partner, I shared a conviction with the team that most generalist investors did not hold: that the next great technology wave would not live on a screen. It would move, sense, and reshape the physical world. That conviction has only sharpened, and the market data is catching up.

Why physical AI is different from the last cycle

Software created enormous value by moving information. Physical AI creates value by moving atoms. The leverage is different, the moats are different, and the founders who succeed are different. When I look at what NVIDIA is building with Alpamayo 2 Super, a 34-billion parameter vision-language-action model trained end-to-end for real-world robotic control, I see the foundation model moment for physical systems arriving in real time.

Alpamayo 2 Super is not a chatbot fine-tuned for robotic arms. It is a new architecture: multimodal reasoning grounded in physical consequence. The model processes camera, lidar, and proprioceptive streams simultaneously, generates action sequences in joint space, and self-corrects from force feedback faster than any prior control stack. NVIDIA trained it on synthetic data at scales that dwarf any real-world robotics dataset, which means it generalizes to manipulation tasks it has never physically executed. That is the inflection.

When foundation models can act, not just predict, every physical system becomes a software opportunity. That is what Alpamayo 2 Super represents.
John Gabriel, IXSAR Capital

Three hardware cycles converging at once

The timing is not coincidental. Robotics, autonomy, and space are each at inflection points simultaneously, and they share an engineering stack. The sensor suites are the same. The edge compute architectures are the same. The perception and planning algorithms are the same. A team that knows how to build a reliable perception stack for an autonomous vehicle can apply that capability to a warehouse robot or a satellite servicing arm.

This is why I invest across Space, Autonomy, and Robotics rather than picking one. The founders building the most interesting companies are not sector specialists. They are systems engineers who understand that the hard problems, reliable real-world operation under uncertainty, are fundamentally the same across all three.

What I look for

I back technical founders who have earned rare domain insight through doing, not through reading about it. The companies I want to back have a defensible engineering moat, a path to real-world deployment rather than just a compelling demo, and a team that understands manufacturing as well as it understands code. The founders of the physical AI era need to think across the entire stack: silicon, firmware, mechanical systems, software, and operations. That full-stack thinking is rare and valuable.

Explore

This site requires JavaScript for the interactive version. For inquiries, email partners@ixsar.com (investors) or companies@ixsar.com (founders).