Chaitanya Lolla
Senior Member Of Technical Staff at AMD
Hyderabad, Telangana, India
Summary
🤩
Rockstar🎓
Top SchoolChaitanya Lolla is a Senior Member of Technical Staff with 10 years of experience building and optimizing deep learning frameworks and inference solutions for AMD GPUs. He blends hands-on framework development in PyTorch, Caffe2 and TensorFlow with low-level GPU work—optimizing FSDP, hipification, and BF16 communication hooks for ROCm. His background spans AI model performance, profiling tools, and custom library ports from his time at Intel/Habana and multiple roles at AMD, giving him a rare full-stack view of training-to-inference pipelines. Earlier experience in J2EE systems and embedded design adds breadth to his engineering judgment and systems thinking. Based in Hyderabad, he frequently contributes to high-profile open-source projects like pytorch/pytorch, where his ROCm fixes and test optimizations improved stability and accuracy. He is driven by curiosity and a habit of prototyping practical tools that turn research ideas into production-ready components.
10 years of coding experience
8 years of employment as a software developer
Master’s Degree Computer Science, Master’s Degree Computer Science at University of North Carolina at CharlotteBPDAVBachelor's Degree Electronics and Communications Engineering, Bachelor's Degree Electronics and Communications Engineering at CVR College of Engineering, Hyderabad
English, Telugu, Hindi
Github Skills (12)
pytorch10
machine-learning10
roc10
gpu10
python10
testing10
fs9
deep-learning9
cuda9
c-language8
devops8
cprogramming-language8
Programming languages (6)
Github contributions (5)
pytorch/pytorch
Dec 2018 - Sep 2020
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
ML Engineer & DevOps Engineer
ML Engineer & DevOps EngineerContributions:6 reviews, 11 commits, 12 PRs in 1 year 10 months
Contributions summary:Chaitanya contributed to the PyTorch repository by implementing and testing features related to ROCm (AMD's GPU platform). They enabled and tested BF16 communication hooks for FSDP on ROCm, fixed circular recursion issues in hipification, and updated ROCm-specific skip decorators for unit tests. The user also optimized FSDP unit tests, limiting world size for improved stability and fixed accuracy issues for hipblasLt for mm use cases on ROCm.
pythongpu-accelerationdeep-learninggpunumpy
lcskrishna/pytorch
Aug 2018 - Jan 2025
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Contributions:301 pushes, 127 branches in 6 years 6 months
pythongpu-accelerationdeep-learninggpuacceleration
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.