BIOGRAPHY
I am currently a PhD student in the Department of Electrical & Computer Engineering at the Stony Brook University. I am also a member of High Performance Data Transfer Tool Group, working with Prof. Dantong Yu and Prof. Shudong Jin. All of my past and current research projects are from the department of CSC at BNL.
I received M.S. from Xidian University in 2009, and B.S. degree in Electrical and Information Engineering from Huazhong University of Science and Technology (HUST) in 2006.
Here is my Resume.
Look at my news and updates at Institute for Advanced Computational Science (IACS), Stony Brook University
________________________________________________________
RESEARCH
My research interest is generally in parallel processing and multicore optimizations. More specifically, I am intrigued by the problems that are related to optimize the end-system to maximize the the end-to-end performance of high-speed Ethernet/InfiniBand networks. The techniques I am usually using include multithreading, asynchronous processing, event-driven design, resource-conscious scheduling, multicore-awareness (NUMA-awareness), and kernel-bypass operations (RDMA, direct I/O, sendfile and etc). I am also interested in developing novel methodology to characterizing, modeling and pinpointing the performance bottleneck in sophisticated parallel/pipelined systems. The tools I am using here include CPU event counters (Intel VTune, OProfile, Linux perf), resource affinity (Linux numactl and libnuma utilities), and cache-coherency protocols (Intel Snoop and AMD HyperTransport). Following are the cool projects that funding my research:
Multicore-Aware Data Transfer Middleware (MDTM) . MDTM is an ongoing project of Fermilab and Brookhaven National Lab, whose goal is to design and build middleware and data transfer application to accelerate data movement at multicore systems.
See our Demo and Poster at IEEE/ACM Supercomputing 2014
FTP100. This project is to design, implement, and evaluate a ultra-high speed data transfer software, RFTP, which take advantage of RDMA and other kernel bypass techniques. I participated in the development and evaluation of RFTP software at 2010, and this project ended at 2013.
________________________________________________________
SOFTWARE
MultiCore-aware Data Transfer (Source code link) MCDT is a novel data transfer tool that adopts I/O-related resource abstraction, capacity-aware thread pools preallocation, storage-centric task mapping and thread-dependent NUMA-aware scheduler to scale up end-to-end data transfer performance over multicore systems. It is proven to be the best-performing TCP/IP-based data transfer tool for multicore servers and 40G/100G LAN/WAN Ethernet networks.
numademo with iomodel A new test module (iomodel) has been extended to linux numademo utility. It takes advantage of memory semantics to simulate I/O operations and model the I/O bandwidth performance for modern high-end NUMA systems. The resulted model can then be used to predict the I/O performance and assist the I/O related resource scheduling.
RFTP RFTP is an RDMA-based data transfer software for high performance networks, such as InfiniBand, RoCE, and iWARP. In the meantime, Linux TCP zero-copy interfaces, sendfile and splice, are also integrated into RFTP to accommodate network environments without RDMA supports. It also integrates an asynchronous I/O module for efficient disk access. The software architecture of RFTP is a combination of event-driven and multi-threaded programming model to maximize network and disk bandwidth utilization and minimize CPU utilization.
________________________________________________________
PUBLICATIONS
Analysis of NUMA Effects in Modern Multicore Systems for the Design of High-performance Data Transfer Applications
Tan Li, Yufei Ren, Dantong Yu, Shudong Jin and Thomas Robertazzi
Future Generation Computer Systems (FGCS), Volume 74, pp 41-50, Sept 2017. [Full Paper]
RAMSYS: Resource-Aware Asynchronous Data Transfer with Multicore SYStems
Tan Li, Yufei Ren, Dantong Yu, Shudong Jin and Thomas Robertazzi
IEEE Transactions on Parallel and Distributed Systems (TPDS), Volume 28, Issue 5, pp 1430-1444, May 2017. [Full Paper]
Resources-conscious Asynchronous High-speed Data Transfer in Multicore Systems: Design, Optimizations, and Evaluation
Tan Li, Yufei Ren, Dantong Yu and Shudong Jin
29th IEEE International Parallel & Distributed Processing Symposium (IPDPS'15), Hyderabad, India, May 2015 (AR: 21.8%) [slides] [Full Paper]
Characterization of Input/Output Bandwidth Performance Models in NUMA Architecture for Data Intensive Applications
Tan Li, Yufei Ren, Dantong Yu, Shudong Jin and Thomas Robertazzi
Proceedings of the 42nd International Conference on Parallel Processing (ICPP'13), Lyon, France, October 2013 (AR: 30%) [slides] [Full Paper]
An Improved Margin Adaptive Subcarrier Allocation with Fairness for Multiuser OFDMA System
Tan Li, Gang Su, Guangxi Zhu, Jun Jiang and Hui Zhang
Proceedings of the 6th International Symposium on Neural Networks (ISNN'09), Wuhan, China, May 2009
Design, Implementation, and Evaluation of a NUMA-Aware Cache for iSCSI Storage Servers
Yufei Ren, Tan Li, Dantong Yu, Shudong Jin and Thomas Robertazzi
IEEE Transactions on Parallel and Distributed Systems (TPDS), vol.PP, no.99, pp.1,1
Design and Performance Evaluation of NUMA-aware RDMA-based End-to-End Data Transfer Systems
Yufei Ren, Tan Li, Dantong Yu, Shudong Jin and Thomas Robertazzi
International Conference for High Performance Computing, Networking, Storage and Analysis (SC'13), Denver, Colorado, Nov 2013
Design and Testbed Evaluation of RDMA-Based Middleware for High-Performance Data Transfer Applications
Yufei Ren, Tan Li, Dantong Yu, Shudong Jin and Thomas Robertazzi
Journal of Systems and Software (JSS), Volume 86, Issue 7, July 2013, Pages 1850-1863
Protocols for Wide-Area Data-Intensive Applications: Design and Performance Issues
Yufei Ren, Tan Li, Dantong Yu, Shudong Jin and Thomas Robertazzi
International Conference for High Performance Computing, Networking, Storage and Analysis (SC'12), Salt Lake City, Utah, November 2012
________________________________________________________
ACTIVITIES
Student member, Institute of Electrical and Electronics Engineers (IEEE)
Student member, IACS, Stony Brook University
Sub-Reviewer, The 23rd International Conference on Computer Communications and Networks, Aug 2014, Shanghai, China
Poster presenter, the International Conference for High Performance Computing, Networking, Storage and Analysis (SC'11), Nov 2011, Seattle, Washington.
Poster presenter & volunteer, the International Conference for High Performance Computing, Networking, Storage and Analysis (SC'10), Nov 2010, New Orleans, Louisiana.
Founder and President, Graduate Badminton Union at Stony Brook, Aug 2013-present
________________________________________________________
AWARDS & HONORS
Stony Brook IACS Travel Award, Stony Brook, NY, 2015
Best Teaching Assistant of Engineering Summer Camp at Stony Brook, Stony Brook, NY, 2011
Merit Student of the University for the top 5% students, HUST, Wuhan, China, 2005
Exempt from Admission Exam, Xidian University, Xi'an, China, 2005
________________________________________________________
COURSES
CSE533 Network Programming. Instructor: Prof. Hussein Badr
AMS540 Linear Programming. Instructor: Prof. Estie M. Arkin
ESE548 Local and Wide Area Network. Instructor: Prof. Milton J. Jackson
ESE550 Network Management and Planning. Instructor: Prof. Dantong Yu
TA:
2009 Fall, 2010 Spring , 2010 Fall and 2014 Spring
ESE 123 Introduction to Electrical & Computer Engineering. Instructor: Prof. David Westerfeld
|