Zhenlin Luo
AI Software Dev. Leader at Intel Corporation
Portland, Oregon Metropolitan Area United States
Summary
👤
Senior🎓
Top SchoolZhenlin Luo is an AI software development leader with nine years at Intel driving high-performance AI and system optimizations across Xeon platforms and memory subsystems. He architects end-to-end AI stacks—spanning MKL integration, quantization algorithms, AMX/AVX-512 tuning, and framework productization in OpenVINO and Intel Neural Compressor—delivering measurable 2x inference gains and enabling multi-node LLaMA2/StarCoder demos. His background in Linux kernel and platform engineering includes accelerating transparent memory compression, shaping Optane persistent memory use cases, and inventing a software-aware tiered memory architecture (patented). A proven open-source contributor, he optimized MKL support in the widely used MXNet project to enhance performance on Intel hardware. Based in Portland, he combines deep systems-level debugging with practical ML model tooling, and is currently building RAG-based private-data chatbots that marry model tuning with platform-specific ISA acceleration.
9 years of coding experience
9 years of employment as a software developer
Bachelor, Mathmatics, Bachelor, Mathmatics at Nankai UniversityMaster, Computer Science, Master, Computer Science at Peking University
English
Github Skills (12)
cpu-architecture10
mxnet10
deep-learning10
intel10
performance-optimization10
architecture10
computer-architecture10
cprogramming-language9
c-language9
convolution9
operator8
batch-normalization8
Programming languages (3)
Github contributions (5)
apache/mxnet
Oct 2016 - Apr 2017
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:
ML Engineer
ML EngineerContributions:10 commits, 16 PRs, 42 comments in 5 months
Contributions summary:Zhenlin primarily contributed to integrating and optimizing Intel Math Kernel Library (MKL) for improved performance within the MXNet deep learning framework. Their work involved adding MKL support for various layers, including convolution, batch normalization, and elementwise operations, as well as addressing related compilation and build issues. The commits demonstrate a focus on enhancing the framework's performance on Intel hardware by leveraging MKL's optimized routines. These contributions included bug fixes and code refinements related to MKL integration.
pythonschedulerdataflowmutationdata-science
zhenlinluo/mshadow
Dec 2016 - Feb 2017
Contributions:4 PRs, 7 pushes, 2 branches in 2 months
cudapytorchcpucppdeep-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.