Hands-on Training Lab: NVIDIA Warp for Computational Physics, GTC 16-19 March 2026

[momo-van]

@shi-eric, @ns5678, and @momo-van will run a hands-on training lab at GTC San Jose, March 16–19, 2026, focused on Warp for Computational Physics.

How to use NVIDIA Warp to Build GPU-Accelerated Computational Physics Simulations

Discover how NVIDIA Warp enables the next generation of GPU-accelerated physics, geometry processing, and differentiable programming. This training lab introduces Warp’s core capabilities and modules, then moves into a hands-on notebook where participants build a high-performance physics solver entirely in Python. This is demonstrated through a practical computational fluid dynamics example based on a 2-D Navier–Stokes formulation. You'll leave with an understanding of what Warp is designed for, how its core building blocks work, and how it can be used alongside deep learning frameworks like PyTorch and JAX in modern computational engineering workflows.

We look forward to seeing the Warp community there. We will publish the material as open-source step-by-step tutorials, product examples, and a technical blog post after GTC.

Footage credit to Autodesk Research, developers of XLB, which supports NVIDIA Warp. [@mehdiataei, et al] are also Warp contributors.

Autodesk.Research.XLB.mp4