Murali Nandhimandalam
GPU Deep Learning SW Engineer at Qualcomm
San Diego, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMurali Nandhimandalam is a GPU deep learning software engineer with over sixteen years of experience building and optimizing ML/DL solutions across mobile, embedded, cloud, and GPU platforms. Currently at AMD after a long tenure at Qualcomm, he specializes in low-level C/C++ and assembly kernel optimization—most recently contributing performance-tuned batch normalization kernels to AMD’s widely used MIOpen library for GFX8 and GFX10 architectures. He pairs a strong DSP and embedded systems background with practical skills in Python, automated build systems, and tooling for debug, calibration, and ROM image workflows. Known for leading cross-functional global teams, he bridges algorithm research and production deployment, often finding hardware-specific optimizations that materially improve runtime behavior.
7 years of coding experience
18 years of employment as a software developer
Masters System science and Automation, Masters System science and Automation at Indian Institute of Science (IISc)BTech Electrical and Electronics Engineering, BTech Electrical and Electronics Engineering at Sri Venkateswara University
English
Github Skills (5)
kernel10
batch-normalization10
performance-optimization10
optimisation10
optimization10
Programming languages (1)
Github contributions (2)
ROCm/MIOpen
Jan 2019 - Jan 2023
AMD's Machine Intelligence Library
Role in this project:
Back-end Developer & Assembly Kernel Engineer
Back-end Developer & Assembly Kernel EngineerContributions:99 reviews, 50 commits, 30 PRs in 4 years
Contributions summary:Murali's primary contribution involves the implementation and optimization of assembly kernels for AMD's Machine Intelligence Library (MIOpen). They added and refined batch normalization assembly kernels, addressing assembler errors and addressing issues for different hardware. These changes included performance optimizations for various batch sizes, GFX8 and GFX10 architectures, and improved the handling of negative variance, demonstrating a focus on kernel performance and hardware-specific adjustments.
amdcluster-computingmachine-learningblasmachine-intelligence
zpwenjh/MIOpen
May 2021 - May 2021
AMD's Machine Intelligence Library
Contributions:1 comment in 1 day
amdmachine-learningaimachine-intelligenceintelligence
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.