Basic Information
Education backgorund：
- Xi’an Jiao Tong University (Bachelor)
- National University of Defense Technology (Master)
Major: Computer science and technology
Birthday：August, 1999 Telephone：182-2994-5747 Email：wrq2021@nudt.edu.cn
Research Areas
- Distributed and Parallel Computing, Collective Communication
- High-performance Networks, including Router Microarchitecture and RDMA NICs.
Educational Experience
2021.9 ~ Present: National University of Defense Technology, Computer Science and Technology, Master Student
- One conference paper on ICS’23 (first author), one patent.
2017.08 ~ 2021.07: Xi’an Jiaotong University, Computer Science and Technology, Bachelor Student
- Xi’an Jiaotong University Scholarship, Outstanding Student of Xi’an Jiaotong University
- National Second Prize in the 2019 Chinese Aircraft Design Challenge
- CET-6 (College English Test band 6) score of 593
Paper Publications
Ruiqi Wang, Dezun Dong, Fei Lei, Ke Wu, Junchao Ma, and Kai Lu. 2023.
Roar: A Router Microarchitecture for In-network Allreduce. In International
Conference on Supercomputing (ICS’23), down load for free.Aoxiang Feng, Dezun Dong, Fei Lei, Junchao Ma, Enda Yu and Ruiqi Wang. 2022. In-network aggregation for data center networks: A survey. In Computer Communications.
Project and Research Experience
1. Accelerator for Distributed AI Training and Parallel Computing
Main Content: Designing an accelerator in a standard router to accelerate the communication bottleneck of distributed AI training and MPI applications: the allreduce operation (collecting data from all hosts and broadcasting the aggregated result to each host), offloading the allreduce operation to the router. This architecture is specifically tailored for RDMA networks.
Achievements: Performance improvement of 2x compared to manually implemented Allreduce operator. Significant reduction in network traffic.
Publication Status: Published at ICS’23 Conference and presented at ChinaSys’23.
2. Fixed-wing Reconnaissance and Strike UAV, in 2019 Chinese Aircraft Design Challenge.
Award: Second National Prize in Chinese Aircraft Design Challenge. In late 2019, the club was featured on China Central Television news channel.
Personal Responsibilities: As the team leader, I was responsible for the development of flight control functions, including:
- Developing flight control software based on NuttX OS and the PX4 open-source framework, allowing the aircraft to automatically aim the camera at the target based on GPS coordinates and perform automated bombing.
- Modifying the source code of the QGroundControl ground station software based on the Qt-5 framework to develop a graphical control interface for automated bombing.
- Integrating the aerodynamic structure and hardware of the aircraft to implement customized control algorithms and programs, ensuring accurate alignment with the flight path during high-speed flight, referencing the PID guidance program of cruise missiles for modified aircraft.
Skills and Personal Characteristics
Programming Languages: Proficient in C/C++, x86-64 assembly, familiar with ARM assembly, Python, able to use Verilog, VHDL, Java, and MySQL.
Hardware Foundation:
- Familiar with high-performance network architecture, including router microarchitecture and RDMA NIC.
- Proficient in x86-64, ARM processor architecture, and assembly language. Experienced in designing multi-stage pipelined processors using Verilog and adept at optimizing program performance by leveraging architectural features.
Software Foundation:
- Proficient in parallel programming, well-versed in MPI and OpenMP programming.
- Knowledgeable in operating system principles, capable of debugging and analyzing the Linux kernel. Familiarity with programming in Linux/Unix environments, concurrent programming, program linking, and related concepts.
- Skilled in debugging and reverse engineering using gdb to analyze source code and vulnerabilities.
Development Tools: Adaptable to any operating system and editor, mainly use Windows and Ubuntu. Frequently used IDEs include VS, VSCode, Anaconda, OMNet++, etc. Experienced in team collaboration using GitHub. Strong learning ability and experience using various frameworks such as openCV, SLAM, PyTorch, Android Studio for development.
Characteristics: A passion for studying computer hardware and software at a low level. Enjoys sports, proactive, resilient under pressure, skilled in communication and team collaboration.
Language Proficiency: Scored 593 on College English Test, Band 6 (CET-6).