Vinci
Category: AI in Manufacturing
Vinci builds a physics AI platform that runs solver-accurate engineering simulation on full-resolution hardware designs, starting with semiconductors. Vinci was founded in 2023. The company is led by Hardik Kabaria. Based in Palo Alto, United States. Total funding raised: $296M. Latest round: Series B. Key investors include Advent, Temasek, Xora Innovation, AMD Ventures, Madrona, Eclipse, Khosla Ventures.
- Founded
- 2023
- Headquarters
- Palo Alto, United States
- Total funding
- $296M
Value proposition
Continuous Physics Reasoning shows engineers how a product will physically behave while it is still being designed. Thermal, thermo-mechanical and convective-fluid results match FEA solver accuracy, run up to 1000x faster, and need no meshing, no approximations and no customer-specific training.
Products and solutions
Continuous Physics Reasoning platform, Foundation Model for Physics, Vinci-Thermal (thermal simulation), thermo-mechanical and warpage analysis, convective fluid simulation, GPU-native physics kernels, agentic orchestration with automated design ingestion of native layout files (OASIS, GDS, IPC-2581).
Unique value
Vinci delivers deterministic, solver-grade physics at manufacturing resolution in seconds to minutes. It works zero-shot on new designs, needs no meshing and no training on customer data, and runs behind customer firewalls.
Target customer
Hardware, semiconductor and electronics engineering teams: chip, package and board designers, and over time automotive, aerospace, memory and advanced-computing hardware makers.
Industries served
Semiconductors and advanced packaging (2.5D/3D IC, chiplets), electronics and PCB. Planned expansion into automotive, aerospace, satellites, memory and advanced computing.
Technology advantage
Pre-trained physics foundation model combined with GPU-native kernels. It handles designs from hundreds of millions to more than 15 billion degrees of freedom. In an EPTC 2025 peer-reviewed study it showed under 2% deviation from commercial solvers at 240x faster runtime. The company says it is deployed on production programs at more than a dozen semiconductor and electronics organizations.
How they differentiate
Unlike surrogate AI models that must be trained or tuned per design or problem class, Vinci runs zero-shot and deterministically at full manufacturing fidelity. Unlike incumbent FEA/EDA tools, it removes meshing and manual setup and runs orders of magnitude faster. This turns simulation from a periodic checkpoint into a continuous step in design.
Main competitors
PhysicsX, Cadence Design Systems, Synopsys (Ansys), BeyondMath
Key partnerships
AMD (strategic investor through AMD Ventures, an AMD SVP gave a supporting quote in the Series B release), EPTC 2025 industry research collaboration, academic advisors from Stanford, UC Riverside and the University of San Francisco.
Notable customers
Undisclosed. The company describes production deployments at leading semiconductor manufacturers but does not name them.
Major milestones
2023: founded by Hardik Kabaria and Sarah Osentoski. December 2025: emerged from stealth with $46M in Seed and Series A funding. December 2025: EPTC 2025 peer-reviewed packaging thermal study. October 2026: $250M Series B at a $1.5B valuation to expand beyond semiconductors.
Growth metrics
$1.5B valuation at the October 2026 Series B. Deployed at more than a dozen semiconductor and electronics organizations. More than half of the top 20 semiconductor companies have benchmarked Vinci.
Market positioning
AI-native engineering simulation and physics-intelligence platform. It positions itself as foundational infrastructure, the next layer after numerical CAE solvers, and expands the EDA/CAE market from a semiconductor beachhead.
Geographic focus
United States headquarters with a global semiconductor customer base; competes with US and UK physics-AI startups and the global EDA/CAE incumbents.
Patents and IP
No patents verified. The company has a peer-reviewed advanced semiconductor packaging thermal study presented at EPTC 2025.
About Hardik Kabaria
Ex-Carbon (ran the software stack); Stanford University PhD, doctoral work on automating high-fidelity meshing for complex geometries in computational geometry and simulation.
Latest news about Vinci
More AI in Manufacturing companies
Official website: https://www.getvinci.ai