科研成果 Publications - 夏 天
出版著作
《片上互连网络:多核/众核处理器关键技术》(任鹏举、夏天译)
ISBN:978-7-121-40344-6,出版社:电子工业出版社,出版日期:2021年1月。
片上互连网络 (豆瓣) (douban.com),片上互连网络――多核/众核处理器关键技术 (京东图书) (jd.com)
期刊论文
[8] Wenzhe Zhao, Guoming Yang, Tian Xia, Fei Chen and Pengju Ren, "HIPU: A Hybrid Intelligenct Processing Unit with Fine-grained ISA for Real-Time Deep Neural Network Inference Applications", IEEE Transactions on Very Large Scale Integration Systems, 2023 (Accepted, CCF B类).
[7] Wenzhe Zhao, Qiwei Dang, Tian Xia, Jingming Zhang, Nanning Zheng and Pengju Ren*, "Optimizing FPGA-based DNN Accelerator with Shared Exponential Floating-Point Format", IEEE Transactions on Circuits and Systems I: Regular Papers, 2023 (Accepted, CCF B类).
[6] 夏天、付格林、曲劭儒、罗中沛、任鹏举,“基于高预测性的稀疏矩阵向量乘法并行计算优化”,计算机研究与发展,2023. (Accepted,中文CCF A类).
[5] Tian Xia, Boran Zhao, Jian Ma, Gelin Fu, Nanning Zheng and Pengju Ren, "An Energy-and-Area-Efficient CNN Accelerator for Universal Powers-of-Two Quantization", IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 70, no. 3, pp. 1242-1255, March 2023. (CCF A类).
[4] Tian Xia, Gelin Fu, Chenyang Li, Zhongpei Luo, Wenzhe Zhao, Nanning Zheng and Pengju Ren, "A Comprehensive Performance Model of Sparse Matrix-Vector Multiplication to Guide Kernel Optimization", IEEE Transactions on Parallel and Distributed Systems, vol. 34, no. 2, pp. 519-534, 1 Feb. 2023.(CCF A类)
[3] Boran Zhao, Tian Xia, Haiming Zhai, Fulun Ma, Yan Du, Hanzhi Chang, Wenzhe Zhao and Pengju Ren, "REMAP: A Spatiotemporal CNN Accelerator Optimization Methodology and Toolkit Thereof", IEEE Transactions on Computer-Aided Design of Integrated Circuit and Systems. 2022. (CCF A类).
[2] Pengchen Zong*, Tian Xia*, Haoran Zhao, Jianming Tong, Zehua Li, Wenzhe Zhao, Nanning Zheng and Pengju Ren, “PIT: Processing-In-Transmission with Fine-Grained Data Manipulation Networks” IEEE Transactions on Computers, 70(6): 877-891, 2021.(共同第一作者,CCF A类)
[1] Tian Xia, Ye Tian, Jean-Christophe Prévotet, and Fabienne NOUVEL. 2019. Ker-ONE: A new hypervisor managing FPGA reconfigurable accelerators. Journal of Systems Architecture 98, C, 453–467, Sept. 2019.(CCF B类)
会议论文
[13] Gelin Fu*, Tian Xia*, Zhongpei Luo, Ruiyang Chen, Wenzhe Zhao and Pengju Ren, "Differential-Matching Prefetcher for Indirect Memory Access", The 30-th IEEE International Symposium on High-Performance Computer Architecture (HPCA), 2024 (Accepted,共同第一作者,CCF A类).
[12] Lili Shi, Haiduo Huang, Bowei Song, Meng Tan, Wenzhe Zhao, Tian Xia and Pengju Ren*, "TAQ: Top-K Attention-aware Quantization For Vision Transformers", 2023 IEEE International Conference on Image Processing (ICIP), 2023 (Accepted, CCF C类).
[11] Gelin Fu*, Tian Xia*, Shaoru Qu, Zhongpei Luo, Shuyu Li, Pengyu Cheng, Runfan Guo, Titong Ding, Pengju Ren, "PrSpMV: An Efficient Predictable Kernel for SpMV", The 41st IEEE International Conference on Computer Design (ICCD), 2023.(Accepted,共同第一作者,CCF B类).
[10] Lingfeng Chen, Tian Xia, Wenzhe Zhao and Pengju Ren, “MI2D: Accelerating Matrix Inversion with 2-Dimensional Tile Manipulations”, The 32th ACM Great Lakes Symposium on VLSI (GLSVLSI),2022, pp 423-429. (CCF C类,指导本科生第一作者)
[9] Chenyang Li, Tian Xia, Wenzhe Zhao, Nanning Zheng and Pengju Ren, “SpV8: Pursuing Optimal Vectorization and Regular Computation Pattern in SpMV”, The 58st Annual ACM/IEEE Design Automation Conference (DAC'58), 2021. pp 661-666. (CCF A类,指导本科生第一作者)
[8] Youzhuo Wang, Wenzhe Zhao, Tian Xia, Nanning Zheng and Pengju Ren, “Joint Critic Mechanism: A Universal Framework for Multi-targets Visual Navigation”, 2021 International Joint Conference on Neural Networks (IJCNN), , Shenzhen, China, 2021, pp. 1-6. (CCF C类)
[7] Tu Zhijun, Jian Ma, Tian Xia, Wenzhe Zhao, Pengju Ren and Nanning Zheng, “CAQ:Context-Aware Quantization via Reinforcement Learning”, 2021 International Joint Conference on Neural Networks (IJCNN), Shenzhen, China, 2021, pp. 1-8.(CCF C类)
[6] Haoran Zhao, Tian Xia, Chenyang Li, Wenzhe Zhao, Nanning Zheng and Pengju Ren*, “Exploring Better Speculation and Data Locality in Sparse Matrix-Vector Multiplication on Intel Xeon” The 38th IEEE International Conference on Computer Design (ICCD), 2020, pp, 601-609. (CCF B类,指导本科生第一作者,校级优秀毕业论文)
[5] Tian Xia, Pengchen Zong, Haoran Zhao, Jianming Tong, Wenzhe Zhao, Nanning Zheng and Pengju Ren* “COCOA: Content-Oriented Configurable Architecture based on Highly-Adaptive Data Transmission Networks” The 30th ACM Great Lakes Symposium on VLSI (GLSVLSI),2020, pp 253-258. (CCF C类)
[4] Tian Xia, Mohamad-Al-Fadl Rihani, Jean-Christophe Prévotet and Fabienne Nouvel, "Ker-ONE: Embedded virtualization approach with dynamic reconfigurable accelerators management," 2016 Conference on Design and Architectures for Signal and Image Processing (DASIP), France, Rennes, 2016, pp. 225-226.
[3] Tian Xia, Jean-Christophe Prévotet and Fabienne Nouvel, "Hypervisor mechanisms to manage FPGA reconfigurable accelerators," 2016 International Conference on Field-Programmable Technology (FPT), Xi'an, China, 2016, pp. 44-52 (CCF C类)
[2] Tian Xia, Jean-Christophe Prévotet and Fabienne Nouvel. “Mini-NOVA: A Lightweight ARM-based Virtualization Microkernel Supporting Dynamic Partial Reconfiguration.” 2015 IEEE International Parallel and Distributed Processing Symposium Workshop (IPDPSW), India, Hydraba, 2015, pp. 71-80.
[1] Tian Xia, Jean-Christophe Prévotet and Fabienne Nouvel. “Microkernel dedicated for dynamic partial reconfiguration on ARM-FPGA platform.” SIGBED Reviews 11, Portugal, Lisbon, 2015, pp. 31-36.
授权专利
[4] 任鹏举,胡康,赵文哲,陈飞,林晓云,楼薇,夏天,郑南宁,“用于通道均衡传输的路由节点”,中国,专利号:ZL202110165948.2,授权公告号:CN113029145B,申请日:2021.02.05,授权日期:2022.08.09.
[3] 任鹏举,张均旺,丁焱,景鑫,赵文哲,夏天,郑南宁, “一种基于强化学习的视觉拓扑导航方法”,中国,专利号:ZL202110165948.2,授权公告号:CN113029145B,申请日:2021.03.01,授权日期:2022.08.09.
[2] 夏天,任鹏举,赵浩然,李泽华,赵文哲,郑南宁,“一种计算架构”,中国,专利号:ZL20201034215.9,授权公告号:CN111522776,申请日:2020.05.13,授权日期:2022.04.05.
[1] Tian Xia, Zhe Liu, "Resource Scheduling Method and Apparatus, Device, and System", US Patent App. 16/984,956, 2020
申请专利
[19] 赵文哲,林晓云,霍志旺,楼薇,张先娆,夏天,任鹏举,“增强神经网络卷积运算数据复用度的高速缓存实现方法”,中国,申请号:202210924541.8,申请日期:2022.08.02
[18] 赵文哲,楼薇,马澄宇,黄静雯,杨国茗,夏天,任鹏举,“一种提高神经网络MAC使用率的多层级联结构”,中国,申请号:2022101138348,申请日期:2022.1.30
[17] 赵文哲,胡康,党祺玮,马澄宇,黄静雯,夏天,任鹏举,“一种加速神经网络计算的多NPU级联结构”,中国,申请号:2022101142485,申请日期:2022.1.30
[16] 任鹏举,党祺玮,陈飞,黄晓帆,张先娆,赵文哲,夏天,“一种高效的神经网络前处理方法”,申请号:202210114249X,申请日期:2022.1.30
[15] 任鹏举,林晓云,赵文哲,党祺玮,刘汉威,夏天,“适合通用硬件电路的神经网络数据存储格式转换方法”,申请号:2022101142502,申请日期:2022.1.30
[14] 赵文哲,杨国茗,夏天,党祺玮,任鹏举,“基于RISC-V的拓展VM指令集编码方法”,申请号:2022101155818,申请日期:2022.1.30
[13] 夏天,刘汉威,马澄宇,党祺玮,赵文哲,任鹏举,“一种基于复杂指令集的RISC-V核及其AI处理器”,申请号:2022101153649,申请日期:2022.1.30
[12] 赵文哲,陈飞,林晓云,胡康,夏天,任鹏举,“基于RISC-V扩展的通用神经网络处理器微架构”,申请号:2022101153615,申请日期:2022.1.30
[11] 任鹏举,陈飞,赵文哲,胡康,杨国茗,夏天,“一种神经网络处理器复杂指令集微架构”,申请号:2022101138140,申请日期:2022.1.30
[10] 任鹏举,陈飞,赵文哲,黄晓帆,杨国茗,夏天,“一种通过算法控制单元进行调度计算的硬件结构”,申请号:202210115362X,申请日期:2022.1.30
[9] 任鹏举,林晓云,霍志旺,楼薇,张先娆,赵文哲,夏天,“一种兼容多维度矩阵乘法的运算单元”,申请号:2022101153102,申请日期:2022.1.30
[8] 任鹏举,林晓云,霍志旺,张先娆,楼薇,赵文哲,夏天,“一种兼容非对称多精度混合乘累加运算的运算单元”,申请号:2022101153653,申请日期:2022.1.30
[7] 赵文哲,胡康,张景铭,付格林,夏天,任鹏举,“一种用于神经网络推理加速的低功耗浮点乘累加运算方法”,申请号:2022101155822,申请日期:2022.1.30
[6] 任鹏举,林晓云,夏天,霍志旺,付格林,赵文哲,“一种卷积运算的硬件自动padding的方法”,申请号:2022101153634,申请日期:2022.1.30
[5] Tian Xia, Linfeng Chen, Wenzhe Zhao, Pengchen Zong, Pengju Ren and Nanning Zheng “A multi-functional data reorganization network”, PCT/CN2021/073039, 申请日:2021.1.21
[4] 任鹏举,李凌阁,丁焱,景鑫,赵文哲,夏天,郑南宁,“一种语义分割辅助的双目视觉无监督深度估计方法”,中国,申请号:202110329765.X,申请日:2021.03.26
[3] 任鹏举,涂志俊,马建,夏天,赵文哲,陈飞,郑南宁,“强化学习自动感知权重分布的卷积神经网络量化方法”,中国,申请号:202110134308.5,申请日:2021.02.01
[2] 夏天,陈凌峰,赵文哲,宗鹏陈,任鹏举,郑南宁, “一种多功能的数据重组网络”,中国,申请号:202011585089.4,申请日:2020.12.28
[1] Tian Xia, Pengju Ren, Haoran Zhao, Zehua Li, Wenzhe Zhao and Nanning Zheng, “A Computer Architecture”, PCT/CN2020/087814.