Christopher Yu

Background

About Christopher Yu

I am a software engineer on the Tools-Simulation team at Pixar. I work primarily on the core physics simulation engine (Fizt), with a particular focus on both explicitly and implicitly integrated constitutive models for strand-based objects like hair. A main concern is in making tools intuitive and art-directable, enabling artists and animators to easily realize the concepts they have in mind. I obtained my PhD in Computer Science from Carnegie Mellon University in August 2021. My thesis work focused on energy-based approaches to non-local collision avoidance on curves and surfaces, and involved building an optimization framework that efficiently reaches self-avoidant configurations using Sobolev gradient descent techniques. We used a wide range of techniques to achieve high performance: numerical techniques like Barnes-Hut approximation, hierarchical matrices, and multigrid solvers; and hardware parallelism using OpenMP and Intel MKL. Art is a major hobby of mine; I mainly work in 2D digital art and 3D modeling.