About Me

Nice to meet you! My name is Minghui. I just graduated from Trinity College (Hartford, CT), where I studied Computer Science and moved to California. Right now I am a software developer in Silicon Valley.

Here are things I like and dislike:

Like Dislike
Coding, Big monitors, Mechanical keyboards, C, Python, Vim, Iced Coffee, Snowboarding, Driving Java, Getting up early, Organize my desk, Trucks and SUVs, Mayonnaise, Carrots

Technical Skills

I am fluent in C and Python. I did a web development internship but right now I mainly do software development on Linux and Python stuff. I love contributing to open source projects. Projects I contribute to include Python, Scikit-learn and the Linux kernel.


I did a web development internship at an insurance software company when I was a junior in college. My responsibilities include designing and building a new website for sharing MS Office templates and maintaining the core product, which was an online insurance management system. After graduation I moved to Silicon Valley. Right now I work on a team that builds a HPC work load manager for clusters.


Name Technology Description
Tin Webserver C, POSIX A multi-threaded web server that supports both static and dynamic contents. It has a one-thread-multiple-active-clients architecture implemented using a thread-pool and handles large number of concurrent requests
Python AES-256 Library Python A python implementation of the Rijndael cipher with 128-bit block size, 256-bit key size and three modes of operations: ECB, CBC, OFB
j3dge Java, OpenGL A Java 3D game engine
Recycalble? iOS, SWIFT An iOS image recognition recycle guide that lets you take a picture of an object and receive its recycle information. Recyclable? won 1st place and best environmental friendly app award at HampHack hackathon
LEMON WEATHER Android, Java, SQLite An Android weather app with beautiful lemon theme
Yet Another Music Player (YAMP) Java, Swing A cross-platform music player with an old-school Winamp like interface
Probabilistic Hypergraph Matching Algorithm on Nvidia GPUs C, CUDA, Thrust, OpenSURF A CUDA implementation of the Probabilistic Hypergraphmatching algorithm on Nvidia GPUs. Achieved 10x+ speedup for large datasets compared to CPU implementation. Presented result at GPU Technology Conference 2015


In college my research interest was on High Performance Computing. I researched Hypergraph matching problem and devised parallel algorithms to solve it efficiently on GPUs.


  1. C.Lin, M.Liu, R.Delany, P.Yoon: Image matching using hypergraphs on GPUs