Lok Koppaka
Senior Hardware Engineer at Microsoft
Bellevue, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolLok Koppaka is a Senior Hardware Engineer with about a decade of hands-on experience designing and optimizing ASIC/FPGA systems for deep learning and memory-intensive applications, currently based in Bellevue, WA. He holds an M.S. in Electrical Engineering and has progressed through roles at Intel, Qualcomm, and multiple positions at Microsoft focused on deep learning system optimization and FPGA acceleration. Proficient in VHDL/Verilog/SystemVerilog, Python, C++, VLSI, and memory architecture, he pairs digital and analog design skills with industry-standard synthesis and verification flows. He has also contributed to the popular DeepSpeed project by implementing low-precision quantization and optimizing transformer kernels, reflecting a bridge between hardware design and ML system performance. Known for turning academic rigor into production-ready optimizations, he seeks roles that deepen verification and testing expertise while delivering measurable system-level gains.
3 years of coding experience
8 years of employment as a software developer
Master of Science (M.S.) Electrical and Electronics Engineering, Master of Science (M.S.) Electrical and Electronics Engineering at Arizona State UniversityBachelor of Technology (B.Tech.) Electronics and Communication engineering, Bachelor of Technology (B.Tech.) Electronics and Communication engineering at R.V.R&J.C COLLEGE OF ENGINEERINGschooling, schooling at Vignan High School,Guntur,India
English, Hindi, Telugu
Github Skills (12)
compression10
pytorch10
machine-learning10
deep-learning10
gpu10
cprogramming-language9
c-language9
inference9
cuda9
data-parallelism8
data-parallel8
parallelization8
Programming languages (1)
Github contributions (3)
deepspeedai/DeepSpeed
Sep 2022 - Jan 2023
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
Role in this project:
ML Engineer
ML EngineerContributions:4 reviews, 67 commits, 19 PRs in 3 months
Contributions summary:Lok contributed to the DeepSpeed library by implementing and optimizing features related to quantization and transformer layers. They introduced a new quantization library with various features, including integer quantization and support for N-dimensional inputs, demonstrating strong expertise in low-precision methods. Additionally, the user refactored and optimized the layer normalization kernel.
billion-parametersfinetuningtrainingmixture-of-expertszero
lokoppakmsft/DeepSpeed
Sep 2022 - Sep 2022
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
Contributions:4 pushes, 2 branches in 1 day
pytorchdeepspeeddeep-learningeffectiveoptimization
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.