Julien Demouth
Sr. Distinguished Engineer (Sr. Director) at NVIDIA
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolJulien Demouth is a Sr. Distinguished Engineer and Sr. Director at NVIDIA with 13 years of industry experience, leading the TensorRT-LLM team and cross-continental engineering groups focused on accelerating deep learning, generative AI, and real-time graphics. His background spans foundational research in computational geometry and computer graphics (PhD and SIGGRAPH/SoCG publications) through GPU-focused R&D at Ubisoft to system-level optimization for GPUs, CPUs and ASICs at NVIDIA. He combines hands-on backend contributions to notable open-source projects—such as integrating CNMeM into Theano—with strategic leadership driving LLM inference, diffusion models, and advanced upscaling/denoising for gaming. Fluent in both research and production environments, he uniquely bridges algorithmic rigor with large-scale deployment and hardware-aware software design. Based in Paris, he manages distributed teams across the US, Europe and Asia while maintaining a deep technical footprint in high-performance computing and rendering.
13 years of coding experience
9 years of employment as a software developer
Ph.D, Computer Science (Computational Geometry), Ph.D, Computer Science (Computational Geometry) at Univ. Nancy / INRIAMS. Computer Science, Algorithmics, MS. Computer Science, Algorithmics at Université Paris 7
French, English
Github Skills (5)
cuda10
c-language10
cprogramming-language10
theano10
gpgpu9
Programming languages (6)
Github contributions (5)
Theano/Theano
Jul 2015 - Jul 2015
Theano was a Python library that allows you to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. It is being continued as PyTensor: www.github.com/pymc-devs/pytensor
Role in this project:
Back-end Developer
Back-end DeveloperContributions:5 commits in 7 days
Contributions summary:Julien primarily contributed to integrating and updating the CNMeM library within the Theano project. Their work involved adapting code to support CNMeM, including modifications to the `cnmem.cpp` and `cuda_ndarray.cu` files. Furthermore, they addressed bug fixes and integrated newer versions of the CNMeM library. This included changes to header files like `cnmem.h` and the core logic within `cnmem.cpp` to align with updates and ensure correct functionality.
mathematical-expressionsmultidimensional-arrayspymcpythontheano
NVIDIA/cnmem
Jul 2015 - Nov 2018
A simple memory manager for CUDA designed to help Deep Learning frameworks manage memory
Contributions:15 commits, 3 PRs, 10 pushes in 3 years 5 months
cudadeep-learning
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.