Ryusuke Sugimoto has received the 2026 Eurographics PhD Award, the 2025 Alain Fournier PhD Dissertation Award, and the SCA Doctoral Dissertation Award, three major honours celebrating excellence in doctoral research. These distinctions recognize his dissertation, Toward General-Purpose Monte Carlo PDE Solvers for Graphics Applications.

Established in 2011, the Eurographics PhD Award recognizes outstanding doctoral research in computer graphics, highlighting exceptional theses and encouraging emerging researchers in the field. The Alain Fournier PhD Dissertation Award is presented annually to honour the best doctoral dissertation in computer graphics completed at a Canadian university and includes a plaque and a $5,000 bursary. The SCA Doctoral Dissertation Award honours a recently graduated doctoral candidate who has successfully defended a dissertation in a field related to computer animation, and recognizes outstanding recent graduates whose work demonstrates strong potential to make impactful and innovative contributions to the field.

“Ryusuke’s dissertation explored how Monte Carlo methods can be used to solve a broad range of partial differential equations relevant to computer graphics,” said Professor Toshiya Hachisuka, who advised Ryusuke with Professor Christopher Batty. “While Monte Carlo approaches have long been successful for rendering problems based on integral equations, his work significantly expands their scope. He developed flexible solvers capable of handling many types of partial differential equations and boundary conditions, truly original work in theory and practice.”

More about this research

Traditional computer simulations of fluid dynamics typically rely on discretizing the Navier–Stokes partial differential equations and solving them using numerical linear algebra techniques. In contrast, Ryusuke developed a Monte Carlo method that fundamentally avoids discretization and linear algebra methods, instead solving partial differential equations directly using computational statistics. This work resulted in a paper at SIGGRAPH Asia 2022 and attracted significant attention from experts in this field, as it introduced the first Monte Carlo method capable of solving the Navier–Stokes equations.

Ryusuke later expanded his research through a project focused on Monte Carlo estimation for boundary-value problems formulated as elliptic partial differential equations. His work was the first to connect Monte Carlo rendering with numerical solutions of boundary-value problems using the walk-on-boundary method. He demonstrated that existing Monte Carlo rendering frameworks, such as GPU-accelerated ray tracing and advanced variance reduction techniques, can be applied directly to solve boundary value problems with high accuracy and efficiency, something previously impossible with conventional discretize-and-solve methods.

This connection was made imperfectly by prior research, and Ryusuke’s work fundamentally outperformed and generalized it over various boundary value problems with different boundary condition types. The research was submitted to SIGGRAPH 2023 as a conference paper, but was subsequently promoted and accepted as a full journal paper instead, an uncommon distinction.

Ryusuke was captivated by this direction of solving partial differential equations using Monte Carlo methods and formed his thesis proposal along this line. He extended the Monte Carlo fluid work from 2022 to support the most common form of the Navier–Stokes equations based on velocity rather than vorticity. He then collaborated with another PhD student to develop a Monte Carlo solver for partial differential equations defined on 3D surfaces and curves, with results accepted as full papers at ACM SIGGRAPH and SIGGRAPH Asia.

In addition to these contributions, Ryusuke’s dissertation includes additional work that extends the applicability of Monte Carlo solvers to a wide range of partial differential equations.