V-Sim
Reaching the impossible of physics simulations
Details
The Founders
Michelle Lu spent more than a decade at NVIDIA, where she served as Director of Simulation Technology. She led the PhysX/Isaac Gym team, the physics engine powering thousands of games and forming the backbone of NVIDIA’s Omniverse and Isaac robotics simulators. With a PhD in Computer Science, Michelle holds several patents in GPU-accelerated simulation and reinforcement learning.
Kier Storey is a distinguished simulation engineer with over 15 years of experience, also from NVIDIA. He was the lead architect on PhysX, contributing innovations in rigid body dynamics, multibody simulations, CFD and FEM. A holder of multiple patents, Kier has been at the frontier of computation physics, pushing performance and fidelity forward.
Together, Michelle and Kier have a rare combination: academic depth, proven technical innovation, and the scars of building the world’s most widely deployed simulation engine. With VSim, they are free to reimagine simulation infrastructure from the ground up.
What Others Missed…
Before VSim, most real-time physics engines traded off accuracy for performance. Approximations were the norm, cloth simulated as stiff meshes, collisions simplified into boxes, and fluids stitched together layer by layer. Robotics and advanced applications were constrained, unable to model complex, nonlinear interactions like flexible materials or fine manipulation tasks.
But Michelle and Kier knew better. Having seen the bottlenecks and technical debt first-hand at NVIDIA, they set out to solve them. VSim’s architecture scales near-perfectly across modern GPUs and cloud clusters, running simulations 50-100x (in some cases 1,000-10,000x) faster than legacy engines
Where others lacked the belief in the second and third order effects of this technology, we saw founders with a deep attachment to a problem and a relentlessness to build something great. Their experience at NVIDIA gave them both the insight into the limits of current tools and the conviction that a new foundation was needed.
The Breakthrough
In 2023, Concept Ventures backed VSim at pre-seed, alongside other investors, to help them build the world’s most flexible, high-performance simulation framework
By 2024, they secured a $20M raise to accelerate research and commercial efforts on the path of advancing this perception layer within Humanoids and generalised robotic use cases. The funding, led by EQT with participation from other global investors, positioned VSim as one of Europe’s most ambitious deeptech bets
Their breakthrough: an engine capable of simulating rigid bodies, deformable bodies, fluids, granular materials, clothing, and robotics - all in real-time, with accuracy once thought impossible.
.jpeg)
