Xianzhi Yu
Researcher at 华为
Haidian District, Beijing, China
Summary
👤
Senior🎓
Top SchoolXianzhi Yu is a researcher with 11 years of experience specializing in high-performance computing and deep learning performance optimization, currently at Huawei Noah Lab in Beijing. He has a strong track record of squeezing performance from heterogeneous systems—optimizing large-scale HPL benchmarks at Sugon, accelerating cryo-EM reconstruction on GPUs, and improving build/install robustness for the high-performance Bolt deep learning library. His work blends low-level systems tuning (CMake, compiler settings, dependency fixes) with algorithmic and pipeline-level improvements across CPU, AMD/NVIDIA GPUs and emerging platforms like Mac M1. A master’s-trained computer scientist, he combines production-grade engineering with research rigor and a knack for resolving platform-specific bottlenecks that often go unnoticed until deployment.
11 years of coding experience
Master's degree, Computer Science, 89/100, Master's degree, Computer Science, 89/100 at University of Chinese Academy of SciencesBachelor's degree, Computer Science and Technology, 91/100, Bachelor's degree, Computer Science and Technology, 91/100 at 山东大学
Github Skills (13)
c-language10
bolt10
cmake10
cprogramming-language10
build-automation10
android9
opencl8
high-performance8
tensorflow7
arm7
tflite7
deep-learning6
deeplearning-ai6
Programming languages (2)
Github contributions (5)
huawei-noah/bolt
Apr 2021 - Oct 2022
Bolt is a deep learning library with high performance and heterogeneous flexibility.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:2 reviews, 57 commits, 27 PRs in 1 year 6 months
Contributions summary:Xianzhi primarily contributed to improving the build and installation process for the Bolt deep learning library. They fixed installation bugs related to dependencies like JSONCPP and OpenCL, updated compiler settings, and modified build scripts. Their work involved modifying CMake files, build configurations, and dependency management, ensuring the library compiled correctly across different platforms and with various dependencies like TensorFlow and TFLite. They also addressed platform-specific build issues, such as those on Android and Mac M1.
mlpcaffe2iosmobiletensorflow
yuxianzhi/darknet-plus
Jun 2019 - Jun 2019
Contributions:5 commits, 2 pushes, 5 branches in 3 days
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.