Pruthvi Madugundu is a seasoned software engineering leader with over a decade of systems and ML-focused experience, currently senior manager of software development at AMD where he leads the AMD PyTorch developer team. He began in embedded DSP and firmware—designing Blu-ray audio engines and low-level codec optimizations—then transitioned to browser engine leadership at Samsung, where he commercialized and optimized Blink for multiple flagship Android devices and pioneered a fingerprint-based web login. At AMD he contributes to PyTorch and ROCm support, improving GPU compatibility, CI, and bfloat16 functionality, blending hands-on engineering with release and DevOps experience on a high-profile open-source ML project. Known for calm, methodical debugging in high-pressure situations, he pairs systems-level rigor with an appetite for new domains, currently pursuing advanced studies in machine learning and computational biology.
8 years of coding experience
20 years of employment as a software developer
BE, Computer Science, BE, Computer Science at Visvesvaraya Technological University
Master's degree, Computer Science, Master's degree, Computer Science at Stony Brook University
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
ML Engineer
Contributions:164 reviews, 37 commits, 66 PRs in 2 years 4 months
Contributions summary:Pruthvi primarily contributed to the ROCm (AMD GPU) support within the PyTorch framework. Their commits focused on addressing compatibility issues related to ROCm, including correcting error code handling for specific APIs and ensuring bfloat16 support. The user also worked on fixing kernel asserts, upgrading the ROCm CI build environment, and enabling and expanding test coverage for ROCm, demonstrating a focus on improving the framework's functionality and stability on AMD hardware. Their work also involved updating Magma for ROCm.
Contributions:4 reviews, 6 commits, 6 PRs in 1 month
Contributions summary:Pruthvi primarily focused on updating the installation instructions within the PyTorch website documentation. Their contributions involved modifying command-line examples, particularly related to installing PyTorch with `pip` and `conda`, across different operating systems, CUDA versions, and ROCm setups. They updated instructions to reflect the correct installation commands for specific configurations, including adjustments for nightly builds and handling different package managers, such as switching to `pip3` for ROCm environments. These updates ensure users can correctly install PyTorch.
pytorchmachine-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.
Request Free Trial
Pruthvi Madugundu - Senior Manager Software Development at AMD