The temporal evolution of physically based fluid simulations is my research topic. I apply machine learning techniques on simulated fluid data and predict the temporal behavior of the physical quantities like pressure or velocity fields.

From 2018 to 2020 I was a research associate in Nils Thuerey’s group.

Besides organizing the Computer Games Laboratory Course I was also supervising Bachelor’s or Master’s Theses. See below for a list of the most recent supervised theses.


E-mail: research (at) steffenwiewel (dot) de
Phone: +
Fax: +
Room: 02.13.060



  • Steffen Wiewel, “Advanced Sequence Learning for Liquid Simulations”, M.Sc. Thesis, TUM, March 2018. (English)
  • Steffen Wiewel, “Practical Applications of the DirectX 11 Depth-Buffer”, B.Sc. Thesis, TUM, September 2015. (German)



Supervised Theses

  • Moritz Becher, “Transfer Learning for Fluid Simulation”, M.Sc. Thesis, TUM, February 2019. (English)
  • Tobias Zengerle, “Physics-based Sequence to Sequence
    Networks with Convolutions”, B.Sc. Thesis, TUM, August 2018. (English)