VTechWorks staff will be away for the Thanksgiving holiday beginning at noon on Wednesday, November 27, through Friday, November 29. We will resume normal operations on Monday, December 2. Thank you for your patience.

Browsing by Author "Zhang, Kai"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Fastmove: A Comprehensive Study of On-Chip DMA and its Demonstration for Accelerating Data Movement in NVM-based Storage Systems
    Li, Jiahao; Su, Jingbo; Chen, Luofan; Li, Cheng; Zhang, Kai; Yang, Liang; Noh, Sam H.; Xu, Yinlong (ACM, 2024)
    Data-intensive applications executing on NVM-based storage systems experience serious bottlenecks when moving data between DRAM and NVM. We advocate for the use of the long-existing but recently neglected on-chip DMA to expedite data movement with three contributions. First, we explore new latency-oriented optimization directions, driven by a comprehensive DMA study, to design a high-performance DMA module, which significantly lowers the I/O size threshold to observe benefits. Second, we propose a new data movement engine, Fastmove, that coordinates the use of the DMA along with the CPU with judicious scheduling and load splitting such that the DMA?s limitations are compensated, and the overall gains are maximized. Finally, with a general kernel-based design, simple APIs, and DAX file system integration, Fastmove allows applications to transparently exploit the DMA and its new features without code change. We run three data-intensive applications MySQL, GraphWalker, and Filebench atop NOVA, ext4-DAX, and XFS-DAX, with standard benchmarks like TPC-C, and popular graph algorithms like PageRank. Across single- and multi-socket settings, compared to the conventional CPU-only NVM accesses, Fastmove introduces to TPC-C with MySQL 1.13-2.16× speedups of peak throughput, reduces the average latency by 17.7-60.8%, and saves 37.1-68.9% CPU usage spent in data movement. It also shortens the execution time of graph algorithms with GraphWalker by 39.7-53.4%, and introduces 1.12-1.27× throughput speedups for Filebench.
  • Thumbnail Image
    Influence of Temperature on the Natural Vibration Characteristics of Simply Supported Reinforced Concrete Beam
    Cai, Yanxia; Zhang, Kai; Ye, Zhoujing; Liu, Chang; Lu, Kaiji; Wang, Linbing (MDPI, 2021-06-21)
    Natural vibration characteristics serve as one of the crucial references for bridge monitoring. However, temperature-induced changes in the natural vibration characteristics of bridge structures may exceed the impact of structural damage, thus causing some interference in damage identification. This study analyzed the influence of temperature on the natural vibration characteristics of simply supported beams, which is the most widely used bridge structure. The theoretical formula for the variation of the natural frequency of simply supported beams with temperature was proposed. The elastic modulus of simply supported beams in the range of −40 °C to 60 °C was acquired by means of the falling ball test and the theoretical formula and was compared with the elastic modulus obtained by the three-point bending test at room temperature (20 °C). In addition, the Midas/Civil finite-element simulation was carried out for the natural frequency of simply supported beams at different temperatures. The results showed that temperature was the main factor causing the variation of the natural frequency of simply supported beams. The linear negative correlation between the natural frequency of simply supported beams and their temperature were observed. The natural frequency of simply supported beams decreased by 0.148% for every 1 °C increase. This research contributed to the further understanding of the natural vibration characteristics of simply supported beams under the influence of temperature so as to provide references for natural frequency monitoring and damage identification of beam bridges.
  • Thumbnail Image
    Propagation topography of redox phase transformations in heterogeneous layered oxide cathode materials
    Mu, Linqin; Yuan, Qingxi; Tian, Chixia; Wei, Chinxi; Zhang, Kai; Liu, Jin; Pianetta, Piero; Doeff, Marca M.; Liu, Yijin; Lin, Feng (Springer Nature, 2018-07-18)
    Redox phase transformations are relevant to a number of metrics pertaining to the electrochemical performance of batteries. These phase transformations deviate from and are more complicated than the conventional theory of phase nucleation and propagation, owing to simultaneous changes of cationic and anionic valence states as well as the polycrystalline nature of battery materials. Herein, we propose an integrative approach of mapping valence states and constructing chemical topographies to investigate the redox phase transformation in polycrystalline layered oxide cathode materials under thermal abuse conditions. We discover that, in addition to the three-dimensional heterogeneous phase transformation, there is a mesoscale evolution of local valence curvatures in valence state topographies. The relative probability of negative and positive local valence curvatures alternates during the layered-to-spinel/rocksalt phase transformation. The implementation of our method can potentially provide a universal approach to study phase transformation behaviors in battery materials and beyond.
  • Thumbnail Image
    Revitalizing the Forgotten On-Chip DMA to Expedite Data Movement in NVM-based Storage Systems
    Su, Jingbo; Li, Jiahao; Chen, Luofan; Li, Cheng; Zhang, Kai; Yang, Liang; Noh, Sam H.; Xu, Yinlong (Usenix Association, 2023)
    Data-intensive applications executing on NVM-based storage systems experience serious bottlenecks when moving data between DRAM and NVM. We advocate for the use of the long-existing but recently neglected on-chip DMA to expedite data movement with three contributions. First, we explore new latency-oriented optimization directions, driven by a comprehensive DMA study, to design a high-performance DMA module, which significantly lowers the I/O size threshold to observe benefits. Second, we propose a new data movement engine, Fastmove, that coordinates the use of the DMA along with the CPU with judicious scheduling and load splitting such that the DMA’s limitations are compensated, and the overall gains are maximized. Finally, with a general kernel-based design, simple APIs, and DAX file system integration, Fastmove allows applications to transparently exploit the DMA and its new features without code change. We run three data-intensive applications MySQL, GraphWalker, and Filebench atop NOVA, ext4-DAX, and XFS-DAX, with standard benchmarks like TPC-C, and popular graph algorithms like PageRank. Across single- and multi-socket settings, compared to the conventional CPU-only NVM accesses, Fastmove introduces to TPC-C with MySQL 1.13-2.16× speedups of peak throughput, reduces the average latency by 17.7-60.8%, and saves 37.1-68.9% CPU usage spent in data movement. It also shortens the execution time of graph algorithms with GraphWalker by 39.7-53.4%, and introduces 1.12-1.27× throughput speedups for Filebench.