Podcasts + Vodcasts
10.6.2026
Co-founder and CEO Hardik Kabaria joined TBPN after Vinci’s $250 million fundraise. He talked about the path from geometry software to a foundation model for physics, why Vinci started with semiconductor thermal analysis, how the company sells into one of the most IP-sensitive industries, and how it plans to expand into new physics modalities.
Watch the Interview
Kabaria traces Vinci’s origins to his work in physics and geometry software. After a PhD at Stanford, he spent eight years at 3D printing company Carbon, rising to VP of software and shipping software to manufacturing floors across Asia, Eastern Europe, the US and Mexico. That work showed him how much software can do for hardware. He founded Vinci three years ago around a simple premise: every part, human-made or natural, is governed by the same universal laws of physics. If a design doesn’t perform in the real world, nothing else about it matters. A team can generate thousands of design ideas, but without a way to grade them against physics, picking the right one is nearly impossible. Today that grading is manual, and Vinci’s foundation model makes it automatic and scalable.
The opportunity spans everything from high-bandwidth memory to satellites, so Vinci deliberately started narrow: thermal problems in semiconductors and electronics. Kabaria says the sale is top-down, through VP-level leaders. The conversation starts with what customers care about, and the answer is consistently time to market and performance. Vinci works on some of the most complex parts in the industry, including HBM, CPUs and GPUs. Their thermal challenges have outgrown existing tools, so a more accurate, higher-fidelity, automated approach earns serious attention.
Semiconductor companies guard their designs closely, and Kabaria explains how Vinci works within that. The platform deploys on-premises, with no forward-deployed engineers and strict data privacy. Vinci also doesn’t need customer designs to train its model. The laws of physics are the same inside a chipmaker’s firewall as outside it, so a model trained on physics works out of the box. Vinci is training on petabytes of data today and plans to scale that by 10x to 100x over the next few quarters. It is expanding from thermal to thermo-mechanical to electromagnetics, with the goal of covering every physics modality that can be described by equations.
"Physics is the compilation layer for hardware."
Hardik Kabaria
Co-Founder and CEO, Vinci
Kabaria’s “compilation layer” framing captures what the platform does in practice. Design generation is accelerating, but physics evaluation has stayed slow and specialist-driven. Vinci turns that evaluation into an automated, scalable step that teams can run continuously, so engineers can confidently choose among many more design options and hit their market windows.
The platform is also built for how the semiconductor industry operates. On-premises deployment, zero forward-deployed engineers, and a model that works without seeing proprietary designs let Vinci meet customers where they are. Because the model is grounded in universal physics rather than any customer’s data, each new modality Vinci adds reaches every deployment, across thermal, mechanical, electromagnetic and future domains.
Vinci is a frontier lab building the foundation model for the physical world. Its deterministic, solver-grounded systems make physics continuously computable inside production engineering workflows and are already running on flagship programs, shifting physics from an episodic simulation bottleneck to continuous infrastructure for design, manufacturing, and reliability decisions.
