Khac Nguyen
Software Engineer at Google DeepMind
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolKhac Nguyen is a software engineer with 11 years of experience specializing in machine learning performance and hardware-aware optimizations, currently working at Google DeepMind to squeeze maximum throughput from TPUs for Gemini models. He has deep expertise across the stack—from high-level quantization and graph transforms to low-level MLIR and instruction-level tuning—built during a multi-year tenure at Google and contributions to TensorFlow's TF Quantizer. Khac’s open-source work includes enabling BatchMatMul support, fixing quantized Gather crashes, and adding Einsum tests, reflecting a focus on robust, production-ready model quantization. Earlier roles in security and networking at Samsung grounded him in systems reliability and platform-level constraints. He holds an MS in Electrical and Computer Engineering from Seoul National University and a bachelor’s degree from Hanoi University of Science and Technology. Notably, Khac blends research-grade rigor with pragmatic engineering to turn model innovations into highly optimized serving pipelines on custom silicon.
10 years of coding experience
9 years of employment as a software developer
Master of Science - MS Electrical and Computer Engineering, Master of Science - MS Electrical and Computer Engineering at Seoul National UniversityBachelor's degree Electrical Electronics and Communications Engineering, Bachelor's degree Electrical Electronics and Communications Engineering at Hanoi University of Science and Technilogy
Github Skills (7)
quantization10
machine-learning10
quants10
deep-learning10
tensorflow10
python9
xla8
Programming languages (7)
Github contributions (5)
tensorflow/tensorflow
Sep 2019 - Jan 2023
An Open Source Machine Learning Framework for Everyone
Role in this project:
ML Engineer
ML EngineerContributions:60 reviews, 301 commits, 8 PRs in 3 years 4 months
Contributions summary:Khac contributed to the TF Quantizer, which is an open-source framework for machine learning. Their commits focused on enhancing the framework by supporting BatchMatMul in the TF Quantizer and fixing a crash that occurred when creating quantized Gather ops with float outputs. They also implemented changes to preserve the output tensor names and improve efficiency. Further enhancements involved supporting batch matmul with broadcastable batch sizes and adding new tests for Einsum, indicating a focus on improving model quantization.
pythondata-sciencedeep-learningmlmachine-learning
mlperf/mobile_app
Nov 2019 - Oct 2020
Mobile App
Contributions:4 reviews, 130 PRs, 188 pushes in 10 months
reactreact-nativeandroidmobile-appnodejs
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.