Miles Macklin
Miles Macklin earlier worked in games and earned a computer science doctorate at Copenhagen on numerical contact methods. He later worked on NVIDIA simulation systems, including PhysX, Flex, and Isaac Gym. In 2021 he co-authored DiSECt, a differentiable simulator for robotic cutting that supports parameter and controller optimisation. He also co-authored Isaac Gym, which keeps physics simulation and policy training on the GPU.
Timeline
-
Earlier career
Numerical physics and interactive systems
Macklin worked in the games industry at Sony, Rocksteady, and LucasArts before NVIDIA. He also earned a computer science doctorate at Copenhagen with Kenny Erleben on numerical methods for contact.
-
2021
Differentiable robotic cutting
Macklin co-authored DiSECt, first submitted in May 2021 and accepted at RSS 2021. It models cutting soft materials and provides derivatives for calibration and controller optimisation. The authors report comparisons with solver and real cutting data.
-
2021
Physics and learning on the GPU
He co-authored Isaac Gym, first submitted in August 2021. The platform runs physics simulation and neural policy training on the GPU. It passes data directly from physics buffers to PyTorch tensors, reducing movement through the CPU during training.
-
As of October 2026
Simulation infrastructure
His research covers simulation for robotics, animation, and interactive applications. His work has been integrated into NVIDIA products including PhysX, Flex, and Isaac Gym.
Sources and credits
Photo: blog.mmacklin.com; Miles Macklin profile photograph; creator not stated. License: Not stated by source. Source: blog.mmacklin.com. Cropped to a square and resized. Ask us to remove or replace this photo.