Negin Sobhani
HPC Consultant - Computational Scientist
Boulder, Colorado, United States
Summary
👤
Senior🎓
Top SchoolNegin Sobhani is an HPC consultant and computational atmospheric scientist in Boulder, Colorado, blending data science, AI, and high-performance computing to advance weather and climate forecasts. With nine years of experience, she has built multi-node, multi-GPU ML workflows (PyTorch DDP/FSDP) on Derecho for MILES CREDIT and led on-prem Kubernetes-based cloud tooling at NCAR. In open source, she contributed to CuPy's GPU-accelerated array operations and performed WRF performance optimizations, including advection routine vectorization and enabling ESMF coupling with CTSM. She has built interactive dashboards and climate tools, such as the LENS-2 Climate Dashboard, to translate ensemble projections for farming communities and enable real-time model-observation comparisons. She leads the NCAR HPC User Group (NHUG) and earned a PhD from the University of Iowa, after foundational studies at the University of Tehran.
10 years of coding experience
9 years of employment as a software developer
Bachelor of Science - BS, Bachelor of Science - BS at University of TehranDoctor of Philosophy (Ph.D.), 3.99/4.00, Doctor of Philosophy (Ph.D.), 3.99/4.00 at University of Iowa
Github Skills (42)
optimizations10
numerical10
python10
model-driven10
data-manipulation10
fortran10
numpy10
model-building10
gpu10
cupy10
numeric10
numericalmethods10
scientific-computing10
modeling10
model-driven-development10
Programming languages (11)
Github contributions (5)
wrf-model/WRF
Dec 2016 - Mar 2021
The official repository for the Weather Research and Forecasting (WRF) model
Role in this project:
Back-end Developer / Performance Engineer
Back-end Developer / Performance EngineerContributions:13 reviews, 5 commits, 3 PRs in 4 years 3 months
Contributions summary:Negin focused on optimizing the performance of the WRF model, specifically targeting computationally intensive sections like the advection limiter loops. Their contributions involved splitting and reordering code within the advection routines to improve vectorization and enhance performance across different compiler environments. Furthermore, the user was involved in enabling the WRF build with the ESMF library to facilitate coupling with CTSM and other external models. They also made modifications to allow CTSM to handle lake points.
forecastingmeteorologyclimateatmospheric-modellingmodel
cupy/cupy
May 2023 - Oct 2023
NumPy & SciPy for GPU
Role in this project:Back-end Developer
Back-end DeveloperContributions:17 reviews, 2 PRs, 25 comments in 5 months
Contributions summary:Negin primarily contributed to the `sliding_window_view` function within the `cupy/cupy` repository, focusing on array manipulation and view creation. They added initial implementation of the sliding window functionality, addressed bug fixes, incorporated documentation and updated tests. The commits demonstrate an understanding of array striding and view creation techniques, crucial for GPU-accelerated numerical computing.
cudapythoncusolvergpunumpy
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.