Eray Inanc
Engineer & Scientist | Microsoft AI for Science | Distributed DL, HPC, Cloud
About me
I'm an engineer and scientist with a PhD in computational sciences, currently working at Microsoft Research AI for Science as a senior research software development engineer. My interests lie on distributed deep learning training, high-performance computing and cloud infrastructure for machine learning workflows.
During my time at the Jülich Supercomputing Centre (JSC), I developed and maintained AI4HPC, an open-source library designed to apply state-of-the-art deep learning models to advanced aerospace use-cases, leveraging HPC systems.
Publications
-
A scientific reasoning model for organic synthesis procedure generation.
Guoqing Liu, Junren Li, Zihan Zhao, E. Inanc, Krzysztof Maziarz, Jose Garrido Torres, Victor Garcia Satorras, Shoko Ueda, Christopher M. Bishop and Marwin Segler, arXiv:2512.13668 (2025). -
Prediction of turbulent boundary layer flow dynamics with Transformers.
R. Sarma, F. Hübenthal, E. Inanc and A. Lintermann, Mathematics 12 (2024). -
Parallel and scalable AI in HPC systems for CFD applications and beyond.
R. Sarma, E. Inanc, M. Aach and A. Lintermann, Frontiers in High Performance Computing 2 (2024). -
Accelerating hyperparameter optimization algorithms with mixed precision.
M. Aach, R. Sarma, E. Inanc, M. Riedel and A. Lintermann, SC-W’23 (2023). -
Large scale performance analysis of distributed deep learning frameworks for convolutional neural networks.
M. Aach, E. Inanc, R. Sarma, M. Riedel and A. Lintermann, J. Big Data 10 (2023). -
Lagrangian filtered density function modelling of a turbulent stratified flame combined with flamelet approach.
S.-J. Baik, E. Inanc, M. Klein and A. Kempf, Phy. Fluids 34 (2022). -
A-posteriori assessment of Large-Eddy Simulation subgrid-closures for momentum and scalar fluxes in a turbulent premixed burner experiment.
L. Engelmann, J. Hasslberger, E. Inanc, M. Klein and A. Kempf, Comp. Fluids 240 (2022). -
An experimental/numerical investigation of non-reacting turbulent flow in a piloted premixed Bunsen burner.
J. Pareja, T. Lipkowicz, E. Inanc, C. D. Carter, A. Kempf and I. Boxx, EoF 63 (2022). -
Scalar gradient and flame propagation statistics of a flame-resolved laboratory-scale turbulent stratified burner simulation.
E. Inanc, A. Kemp and N. Chakraborty, C&F 238 (2022). -
Numerical simulation of pulsed and stratified combustion.
E. Inanc, PhD Thesis, Universität Duisburg-Essen (2021). -
Fully Resolved Auto-Igniting Transient Jet Flame Simulation.
E. Inanc and A. Kempf, High Performance Computing in Science and Engineering ’19 249-262 (2021). -
Effect of sub-grid wrinkling factor modelling on the Large Eddy Simulation of turbulent stratified combustion.
E. Inanc, A. Kempf and N. Chakraborty, CTM 25 (2021). -
A simple post-processing method to correct species predictions in artificially thickened turbulent flames.
P. Gruhlke, E. Inanc, R. Mercier, B. Fiorina and A. Kempf, Proc. Comb. Inst. 38 (2020). -
Detailed simulations of the DLR auto-igniting pulsed jet experiment.
E. Inanc, J. T. Lipkowicz and A. Kempf, Fuel 284 (2020). -
Analysis of mixture stratification effects on unstrained laminar flames.
E. Inanc, N. Chakraborty and A. Kempf, C&F 219 (2020). -
Numerical study of a pulsed auto-igniting jet flame with detailed tabulated chemistry.
E. Inanc and A. Kempf, Fuel 252 (2019). -
Studying transient jet flames by high-resolution LES using premixed flamelet chemistry.
E. Inanc, F. Proch and A. Kempf, DLES XI:237-243 (2019). -
High-resolution LES of a starting jet.
E. Inanc, M. T. Nguyen, S. A. Kaiser and A. Kempf, Comp. Fluids 140 (2016).
Talks & Posters
Selected talks
- Reconstruction of near-wall quantities from filtered flows using Convolutional Defiltering Model, ParCFD (2023), Cuenca/Ecuador
- Parallel and scalable deep learning to reconstruct actuated turbulent boundary layer flows. Part II: autoencoder training on HPC systems, ParCFD (2022), Alba/Italy
- Fully resolved auto-igniting transient jet flame simulation, 22. Results and Review Workshop of the HLRS (2019), Stuttgart/Germany
- Numerical study on pulsed jet flames: tabulated methods against direct chemistry, 29. Deutscher Flammentag (2019), Bochum/Germany
- LES of a pulsating jet flame in a hot co-flow, ETMM 12 (2018), Montpellier/France
- Statistical analysis of pulsating methane flames issued into hot co-flow by LES with FGM, JMGICS (2018), Sorrento/Italy
- Investigation of non-premixed piloted methane flames by LES with flamelet generated manifolds, MCS 10 (2017), Naples/Italy
- Studying transient jet flames by high-resolution LES using premixed flamelet chemistry, DLES 11 (2017), Pisa/Italy
Posters
- AI4HPC: library to train AI models on HPC Systems using CFD datasets, NeurIPS 23 (2023), New Orleans/USA
- Large-Eddy Simulation of auto-igniting pulsating methane flames with multi-dimensional tabulated chemistry, Combustion Symposium 37 (2018), Dublin/Ireland
- Combustion regime investigations of turbulent flames by flame-resolved simulations, TNF 14 (2018), Dublin/Ireland
- Large Eddy Simulation of pulsating jet flames with detailed tabulated chemistry, 28. Deutscher Flammentag (2017), Darmstadt/Germany
- Investigation of the auto-ignition of an impulsively started methane jet emitting into a vitiated co-flow, Future Fuels Workshop (2016), Thuwal/Saudi-Arabia
Awards
- Gauss GCS Large Scale Project Award: 48 million CPUh in High-Performance Computing Center Stuttgart (HazelHen, Cray XC40): 2018 – 2019 (Kz: 44141 GCS-JFLA)
Media