Flexcompute Raises Funding From Samsung Ventures and Applied Ventures
Flexcompute, a company that builds physics-based simulation software, has raised strategic investment from Samsung Ventures and Applied Ventures, the venture arm of Applied Materials, among other investors. The money will go toward its Physics World Model, an AI system meant to predict how designs behave in the real world. Financial details were not disclosed.
- Flexcompute raised strategic investment from Samsung Ventures and Applied Ventures, among other investors; amounts were not disclosed.
- The money will fund its Physics World Model, which combines neural networks with first-principles physics via Tokenized Physics.
- Flexcompute says it solved a 5-billion DOF package warpage model within 2% of commercial finite-element results in minutes.
- Customers access the model today through Physics Twins, used by engineers and AI agents.
- Funds will go to research hiring, GPU compute and expansion across fluids, electromagnetics, structures, heat and multiphysics.
What Flexcompute is building
According to Flexcompute, the Physics World Model is AI that predicts how any design will behave in the real world by learning the laws of physics themselves. The company gives examples ranging from a chip package warping under heat to a spacecraft reentering the atmosphere.
The company contrasts its method with current AI, which it says learns from the past: text, images and simulation results. Flexcompute's team of more than 70 PhDs has built CFD, electromagnetic, structural and thermal solvers from the ground up on GPUs. Through an approach it calls Tokenized Physics, the company says it builds both neural networks and first-principles physics directly into the model. In its description, every prediction obeys the same conservation laws as the physical world and stays accurate on designs the model has never seen.
A semiconductor packaging test case
Flexcompute points to a semiconductor example as evidence that the approach works on problems traditional simulation cannot reach. For a full advanced semiconductor package with more than 800,000 solder joints, an ASIC and 32 optical engines, the company says it solved a 5-billion degree-of-freedom (DOF) warpage model within 2% of commercial finite-element results in minutes.
Yu Cao, Chief AI Officer of Applied Materials, said the limit for industries applying AI to engineering is "how quickly each one can be validated against the laws of physics." He added that Flexcompute has built "simulation fast enough to keep pace with AI and rigorous enough for production decisions," which in advanced manufacturing means more of the design cycle happens digitally, saving time and cost.
A representative of Samsung Ventures said the firm believes the Physics World Model "will shape how the next generation of hardware is developed."
How customers use it and where the money goes
Customers can access the Physics World Model today through Physics Twins. These are models of their own products that engineers and AI agents can question, test and optimize before anything is built. Under this workflow, engineers set the objective and approve the result, agents run the iterations, and physics governs what works.
Vera Yang, Co-Founder and President of Flexcompute, described the goal as building "the verification layer between bits and atoms." She said any engineer or AI agent could ask the model whether a design will work before it is built.
Flexcompute plans to use the investment to grow its research team, scale GPU compute, and extend the Physics World Model across fluids, electromagnetics, structures, heat, and multiphysics couplings.
What it means for hardware design
The backing from a semiconductor equipment supplier and a major electronics group's venture arm points to interest from hardware makers in faster, physics-grounded verification. Flexcompute says engineering teams at more than 250 companies across aerospace, automotive, semiconductors, photonics and advanced manufacturing already use its tools to evaluate, explore and optimize designs. The company was founded by researchers from Stanford and MIT.
The performance claims, including the warpage result, come from Flexcompute itself and the announcement did not include independent validation or the size of the funding round.
