Hi! This is my resume

Education

Ph.D. Student in Electrical and Computer Engineering, Duke University, 2023 - Now
M.S. in Computer Science and Information Engineering, National Cheng Kung University, 2022
B.E. in Computer Engineering, University of Hong Kong, 2018

Sep 2023 - Now: Research Assistant, Electrical and Computer Engineering, Duke University
- Supervisor: Prof. Tingjun Chen
- Design a real-time wideband RF system based on Xilinx’s RFSoC platform [WiNTECH’24]
Feb 2021 - Jul 2023: Research Assistant, Institute of Information Science, Academia Sinica
- Supervisor: Prof. Yuan-Hao Chang
- Graph processing on dual-addressing memory [ICCAD’22]
Sep 2020 - Jun 2022: Research Assistant, Computer Architecture & IC Design Lab, NCKU
- Supervisor: Prof. Ing-Chao Lin
- CNN accelerator with CLIP-Q network quantization on FPGA [TCAS-I]
Summer 2016: Intership, TCL Corporate Research (HK) Co., Ltd
- Assist researchers on Structure from Motion, SLAM algorithms

2022: IEEE Taipei Section Best Dissertation Award
2022: IEEE Tainan Section Best Master Thesis Award
2022: Master Thesis Award from Institute of Information and Computing Machinery (IICM)
2014 - 2018: HKU Foundation Scholarship for International Students
2016 - 2017: Reaching Out Award scholarship from HKSAR gov.
2018: IEEE Hong Kong Final Year Project Merit Award

SPEAR: A 1.25 GHz bandwidth real-time RF testbed
- SPEAR is an SDR platform based on the Xilinx RFSoC ZCU216 evaluation board capable of supporting real-time streaming of signals with a bandwidth of up to 1.25 GHz employing the direct RF radio architecture.
- It comes with an DSP pipeline written in Python and reaches the 3GPP EVM requirement of 3.5% with 256QAM modulation.
- Github: https://github.com/functions-lab/SPEAR
A SW/HW co-designed RISC-V CNN accelerator for mask detection
- RISC-V core consists of: pipelined RV32I core, I-cache/D-cache, AXI bus, DMA, DRAM/ROM controller, Interrupt manager. It also handles the booting sequence, data movements, the control of acceleration unit, and system interrupts.
- Apply network compression and quantization on a mask detection CNN model.
- Inference the compressed CNN model on an HW acceleration unit.
- Github: https://github.com/WeiCheng14159/VSD_CNN_accelerator
Contribute to ria-jit (an open source RISC-V to x86 binary translator)
- Find and fix a divide by zero bug with RISC-V compliance tests.
Contribute to srv32 (an open source 3-stage pipeline RV32IM core)
- Verify and contribute RV32C instructions to the existing implementation.

Cell-based Digital IC tapeout, Taiwan Semiconductor Research Institute
- Design, tapeout, and verification of an UMC 0.18 um process chip
Class representative for Computer Engineering major students during 2015-2016, 2017-2018