曹伟,博士,现任复旦大学大数据研究院副研究员。2006年毕业于哈尔滨工业大学,获工学博士,2006年9月至2009年6月在复旦大学电子科学与技术博士后流动站从事博士后科研工作。主要研究方向是领域专用软硬件协同计算、可重构计算、领域专用硬件加速器结构和可编程逻辑器件结构。2015年获上海市技术发明三等奖一项(排名第四)。迄今,已发表SCI和EI论文30余篇,第一发明人授权发明专利3项。承担十四五国家重点研发计划课题和子课题各1项。
主持科研项目列表:
[1].国家重点研发计划“先进计算与新兴软件”专项,“动态可伸缩的拟态计算系统及构建方法”项目,课题3“拟态计算系统全栈式工具链研究”,执行周期:2023.03~2026.02,国拨经费:1084万
[2].国家重点研发计划“微纳电子专项”,“晶圆级芯片系统级开发环境设计与验证研究”项目, 课题2“晶圆级芯片高效编译优化技术”-子课题“异构平台代码生成”,执行周期:2022.11-2025.10,国拨经费:100万
代表性论文列表:
1.期刊论文:
[1].Xitian Fan, Guangwei Xie, Zhongcheng Huang, Wei Cao*(曹伟,通讯作者) , LingliWang:Acceleration of Rotated Object Detection on FPGA. IEEE Trans. Circuits Syst. II Express Briefs 69(4): 2296-2300 (2022) (SCI,2022年影响因子/JCR分区:3.691/Q2 )
[2].Siyu Xiong, Guoqing Wu, Xitian Fan, Xuan Feng, Zhongcheng Huang, Wei Cao1*(曹伟,通讯作者) , Xuegong Zhou, Shijin Ding,Jinhua Yu, Lingli Wang and Zhifeng Shi, MRI-based brain tumor segmentation using FPGA-accelerated neural network,BMC Bioinformatics,(2021) 22:421(SCI,2022年影响因子/JCR分区:3.307/Q2 )
[3].Di Wu, Xitian Fan, Wei Cao*(曹伟,通讯作者) and Lingli Wang,SWM:A High-Performance Sparse-Winograd Matrix Multiplication CNN Accelerator, IEEE Transactionson Very Large Scale Integration (VLSI) Systems Print ISSN: 1063-8210 Online ISSN: 1557-9999Digital Object Identifier: 10.1109/TVLSI.2021.3060041,(2021)(SCI,2022年影响因
子/JCR分区:2.775/Q2 )
[4].Xitian Fan , Di Wu, Wei Cao*(曹伟,通讯作者), Wayne Luk and Lingli Wang,Stream Processing Dual-Track CGRA for Object Inference,IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS, VOL. 26, NO. 6, JUNE 2018 (SCI,2022年影响因子/JCR分区:2.775/Q2)
2.会议论文:
[1].Jingbo Gao, Yu Qian, Yihan Hu, Xitian Fan, Wai-Shing Luk, Wei Cao*(曹伟,通讯作者) , Lingli Wang: LETA: A lightweight exchangeable-track accelerator for efficientnet based on FPGA. 2021 International Conference on Field-Programmable Technology (FPT2021,可重构计算和FPGA领域四大重要国际会议之一)
[2].Xuan Feng, Yue Li, Yu Qian, Jingbo Gao, Wei Cao*(曹伟,通讯作者) , Lingli Wang: A High-Precision Flexible Symmetry-Aware Architecture for Element-Wise Activation Functions. 2021 International Conference on Field-Programmable Technology (FPT 2021,可重构计算和FPGA领域四大重要国际会议之一)
[3].Yue Li, Wei Cao*(曹伟,通讯作者), Xuegong Zhou* and Lingli Wang, A Low-Cost Reconfigurable Nonlinear Core for Embedded DNN Applications, 2020 InternationalConference on Field-Programmable Technology (FPT2020,可重构计算和FPGA领域四大重要国际会议之一)
[4].Mingjun Jiao, Yue Li, Pengbo Dang, Wei Cao*(曹伟,通讯作者) and Lingli Wang, A High Performance FPGA-based Accelerator Design for End-to-End Speaker Recognition System, 2019 International Conference on Field-Programmable Technology (FPT2019,可重构计算和FPGA领域四大重要国际会议之一)
[5].Di Wu, Wei Cao*(曹伟,通讯作者) and Lingli Wang, SpMM-WCNN: A Novel Highperformance Sparse Matrix-matrix Multiplication Accelerator for Winograd CNN Inference, 2019 International Conference on Field-Programmable Technology (FPT2019,可重构计算和FPGA领域四大重要国际会议之一)
[5].Li Jiao, Cheng Luo, Wei Cao*(曹伟,通讯作者), Xuegong Zhou, Lingli Wang,Accelerating Low Bit-Width Convolutional Neural Networks With Embedded FPGA,27th IEEE International Conference on Field-Programmable Logic and Applications, Sep.2017(FPL2017,可重构计算和FPGA领域四大重要国际会议之一)
[6].Huimin Li, Xitian Fan, Li Jiao, Wei Cao(曹伟,通讯作者), Xuegong Zhou, Lingli Wang, A High Performance FPGA-Based Accelerator for Large-Scale Convolutional Neural Networks,26th IEEE International Conference on Field-Programmable Logic and Applications, Switzerland, pp.69-77, August 2016 (FPL2016,可重构计算和FPGA领域四大重要国际会议之一)
[7].X Fan,H Li,Wei Cao*(曹伟,通讯作者),Lingli Wang,DT-CGRA: Dual-Track Coarse-Grained Reconfigurable Architecture for Stream Applications,26th IEEE International Conference on Field-Programmable Logic and Applications, Switzerland, August 2016 (FPL2016,可重构计算和FPGA领域四大重要国际会议之一)
专利:
1.曹伟(第一发明人),王伶俐等,“一种低位宽卷积神经网络可重构计算单元”,专利号:ZL 2018 1 0318783.6,2021年授权
2.曹伟(第一发明人),王伶俐等,“一种可重构卷积神经网路的硬件互连系统”,专利号:ZL 2018 1 0358433.6,2021年授权
3.曹伟(第一发明人),王伶俐等,“一种基于LBP特征的人脸识别硬件架构”,专利号:CN201510688167.6,2018年授权
教学:
研究生课程“可重构计算”(INFO630077),秋季学期(2011至今),3学分。
联系方式:
电子邮箱:caow@fudan.edu.cn
办公室:复旦大学邯郸校区计算中心B305