Zhengyang Liu

Software Development Engineer
Amazon Web Services
Last seen on Stack Overflow on Sep 20, 2015


Tech you dislike

Experience show all

Software Development Engineer
Amazon Web Services

January 2014 – Current

Elastic Beanstalk

System Developer
BNP Paribas

June 2009 – June 2010

building a in-house trading amendment management system with Intalio's BPMS suite for the Hong Kong branch of the world's biggest bank

Independent iOS Developer

2010 – 2012


MCS Computer Science
University of Virginia

2011 – 2013

BSc in Computing
The Hong Kong Polytechnic University

2006 – 2011

Open Source show all

GitHub, Oct 2013 - Dec 2013

A Mach-O loader for FreeBSD


GitHub, Dec 2011; followed by 9 people; forked 6 times

OpenCL support for OpenSSL (currently only AES is implemented)


GitHub, Feb 2013 - Jan 2014

display currently playing item on status bar

GitHub, Feb 2013; followed by 3 people

Do Not Disturb Toggle for SBSettings

GitHub, Feb 2013; forked 2 times

Force legacy apps to launch in full screen on iPhone 5.

Writing show all

A minimal Mach-o x64 executable for OS X

A 288-byte Mach-o executable

A Minimal (168-byte) Mach-o ARM executable for iOS

A 168-byte Mach-o ARM executable for iOS

A minimal Mach-o x64 executable for OS X Mountain Lion

A minimal Mach-o x64 executable for OS X Mountain Lion.

On Using Virtual Circuits for GridFTP Transfers (Supercomputing 12)

GridFTP transfer logs obtained from NERSC, SLAC, and NCAR, were analyzed. The goal of the analyses is to characterize these transfers and determine the suitability of dynamic virtual circuit (VC) service for these transfers instead of the currently used IP-routed service. Given VC setup overhead, the first analysis of the GridFTP transfer logs characterizes the duration of sessions. Of the NCAR-NICS sessions analyzed, 56% of sessions would have been long enough to be served with dynamic VC service. An analysis of transfer throughput across four paths, NCAR-NICS, SLAC-BNL, NERSC-ORNL and NERSC-ANL, shows significant variance. An analysis of the potential causes of this variance shows that server-related factors are more important than network-related factors. This is because most of the network links are lightly loaded, which implies that throughput variance is likely to remain unchanged with virtual circuits.