Alum Wins SIGGRAPH Dissertation Award Friday, June 6, 2025 - by Adam Kohlhaas SCS alum Rohan Sawhney received the 2025 SIGGRAPH Outstanding Doctoral Dissertation Award for research that redefines how computers simulate and process geometric data.Rohan Sawhney, an alumnus of Carnegie Mellon University’s School of Computer Science, has received the 2025 SIGGRAPH Outstanding Doctoral Dissertation Award for research that redefines how computers simulate and process geometric data. His dissertation introduces a novel framework that moves beyond traditional finite element methods by applying grid-free Monte Carlo techniques to solve fundamental partial differential equations.Presented annually at the SIGGRAPH Conference, the award recognizes recent doctoral graduates whose work advances the fields of computer graphics and interactive techniques. Sawhney completed his Ph.D. in computer science at CMU with adviser Keenan Crane, the Michael B. Donohue Associate Professor of Computer Science and Robotics.Now a senior research scientist at NVIDIA, Sawhney’s work draws inspiration from Monte Carlo rendering methods, the crown jewel of compute graphics long used for photorealistic image generation, to develop robust simulation tools that do not require conventional meshing or discretization.In one key contribution, Sawhney extends the Walk on Spheres (WoS) algorithm to handle partial differential equations with spatially varying coefficients — a major challenge in modeling complex, heterogeneous materials. Originally developed to solve Laplace’s equation, WoS is a probabilistic algorithm that simulates Brownian motion by sampling only the exit points from a sequence of nested spheres, rather than tracking the full path of the motion. This enables millions of such trajectories to be simulated in parallel to better utilize the capabilities of modern multi-core computers compared to traditional grid-based methods. Sawhney’s version draws on techniques from volumetric rendering, particularly null-scattering, to develop unbiased Monte Carlo estimators that solve these equations accurately and efficiently — without meshing — even on highly detailed geometries. "I was inspired by the culture CMU fosters in its Ph.D. program — one that emphasizes tackling hard problems and pursuing fundamental research," Sawhney said. "My favorite part of the experience was collaborating with faculty and students who fully embraced this spirit."For more information about the SIGGRAPH award, visit their website. Detailed information about Sawhney’s research can be found on his website.Media ContactAaron Aupperlee | 412-268-9068 | aaupperlee@cmu.edu