Zhengchun Liu (刘正春)

Postdoctoral Appointee at Argonne National Laboratory

Home Research Publication News Resume

Robust Analytic Models for Science at Extreme Scales

At the Mathematics and Computer Science in the Argonne National Laboratory. He is developing end-to-end analytical performance models to transform understanding of the behavior of science workflows in extreme-scale science environments. These models are developed to predict the behavior of a science workflow before it is implemented, to explain why performance does not meet design goals, and to architect science environments to meet workflow needs.

He is focused on:

  • Modeling and simulating end-to-end data transfers over wide-area network;
  • Analyzing www.globus.org transfer log towards explaining the wide-area network data transfer performance;
  • Building modeling and simulation program that can effectively and efficiently explain the behavior of scientific workflows over a distributed infrastructure.

High Performance Computing and Simulation

At the Computational Sciences and Engineering Division in the Oak Ridge National Laboratory. He worked on:

  • A framework for efficient simulation on multi-GPU and multi-Core clusters, it is designed to accommodate the hierarchical organization as well as heterogeneity of current state-of-the-art parallel computing platforms. We design it to support agent-based simulation and 3D finite difference method based simulation (stencil computation);
  • Interactive, graphical processing unit-based evaluation (faster than real time) of evacuation scenarios at the state scale; [code & demo ]
  • Implemented an earthquake wave propagation model on multiple GPUs using CUDA and the framework described in (1). [code & demo ]

Agent-Based Model and Simulation of Emergency Department

His PhD thesis entitled: Modeling and Simulation for Healthcare Operations Management using High Performance Computing and Agent-Based Model, supervised by Emilio Luque, is about high performance computing based simulation for the decision support of healthcare system operations management. Specifically, simulating the Emergency Department (ED) by using agent-based modeling techniques and making it work as a part of decision support system. I accomplished the modeling, implementation, calibration and validation work.

Since ED is a typical complex system, agent-based simulation technique was used to model the ED directly from an individual level, i.e., the behavior of staff, physical resources and patients. The system-level behavior, that of the system as a whole was considered to be emerged from these individual level behavior. Through this way, the model can represent more detailed information from bottom-up and capable to identify root causes of problems from individual behavior level.

The model has been verified and validated for an ED in Catalunia, Spain. High Performance Computing technique was used to simulate multi-scenarios simultaneously, and optimize unknown model parameters under data scarcity. By this means, the simulator can execute a large number of simulation scenarios in an acceptable period.

Automatic Model-parameter Calibration

One of the key issues in calibration is the acquisition of valid source information from the target system. We developed an automatic calibration (tuning) tool that is released with the general emergency department model. This tool enables the simulation users to calibrate model parameters for their own emergency department system without the involvement of model developers. They believe that the tool is promising for promoting the application of simulation on emergency department related studies.

Embedded Technique

In addition, he is also particularly interested in embedded system. He worked as an embedded engineer for three years in industrial area. He is experienced on both hardware design and embedded software (firmware) developing. He is very optimistic about the application of embedded devices on high performance computing, such as FPGA and DSP for specific speedup.

Machine Learning

He is quite enthusiastic about using artificial intelligence techniques to solve practical problems and make our life better. He keeps learning AI related advances in his spare time solely driven by his interests. Recently, they used machine learning algorithms to explain / predict wide area data transfer performance. He also tries to use deep reinforcement learning to achieve a smart data transfer node, he consider it as the first step to smart HPCC.


Prototype before polishing. Get it working before you optimize it.