LIVIVO - The Search Portal for Life Sciences

zur deutschen Oberfläche wechseln
Advanced search

Search results

Result 1 - 5 of total 5

Search options

  1. Book ; Online: Performance and Power Modeling and Prediction Using MuMMI and Ten Machine Learning Methods

    Wu, Xingfu / Taylor, Valerie / Lan, Zhiling

    2020  

    Abstract: In this paper, we use modeling and prediction tool MuMMI (Multiple Metrics Modeling Infrastructure) and ten machine learning methods to model and predict performance and power and compare their prediction error rates. We use a fault-tolerant linear ... ...

    Abstract In this paper, we use modeling and prediction tool MuMMI (Multiple Metrics Modeling Infrastructure) and ten machine learning methods to model and predict performance and power and compare their prediction error rates. We use a fault-tolerant linear algebra code and a fault-tolerant heat distribution code to conduct our modeling and prediction study on the Cray XC40 Theta and IBM BG/Q Mira at Argonne National Laboratory and the Intel Haswell cluster Shepard at Sandia National Laboratories. Our experiment results show that the prediction error rates in performance and power using MuMMI are less than 10% for most cases. Based on the models for runtime, node power, CPU power, and memory power, we identify the most significant performance counters for potential optimization efforts associated with the application characteristics and the target architectures, and we predict theoretical outcomes of the potential optimizations. When we compare the prediction accuracy using MuMMI with that using 10 machine learning methods, we observe that MuMMI not only results in more accurate prediction in both performance and power but also presents how performance counters impact the performance and power models. This provides some insights about how to fine-tune the applications and/or systems for energy efficiency.

    Comment: to be published in Cray User Group Conference 2020
    Keywords Computer Science - Machine Learning ; Computer Science - Performance
    Publishing date 2020-11-12
    Publishing country us
    Document type Book ; Online
    Database BASE - Bielefeld Academic Search Engine (life sciences selection)

    More links

    Kategorien

  2. Book ; Online: Hybrid Workload Scheduling on HPC Systems

    Fan, Yuping / Rich, Paul / Allcock, William / Papka, Michael / Lan, Zhiling

    2021  

    Abstract: Traditionally, on-demand, rigid, and malleable applications have been scheduled and executed on separate systems. The ever-growing workload demands and rapidly developing HPC infrastructure trigger the interest of converging these applications on a ... ...

    Abstract Traditionally, on-demand, rigid, and malleable applications have been scheduled and executed on separate systems. The ever-growing workload demands and rapidly developing HPC infrastructure trigger the interest of converging these applications on a single HPC system. Although allocating the hybrid workloads within one system could potentially improve system efficiency, it is difficult to balance the tradeoff between the responsiveness of on-demand requests, the incentive for malleable jobs, and the performance of rigid applications. In this study, we present several scheduling mechanisms to address the issues involved in co-scheduling on-demand, rigid, and malleable jobs on a single HPC system. We extensively evaluate and compare their performance under various configurations and workloads. Our experimental results show that our proposed mechanisms are capable of serving on-demand workloads with minimal delay, offering incentives for declaring malleability, and improving system performance.
    Keywords Computer Science - Distributed ; Parallel ; and Cluster Computing
    Subject code 000
    Publishing date 2021-09-11
    Publishing country us
    Document type Book ; Online
    Database BASE - Bielefeld Academic Search Engine (life sciences selection)

    More links

    Kategorien

  3. Article ; Online: Detailed analysis of BALB/c mice challenged with wild type rotavirus EDIM provide an alternative for infection model of rotavirus.

    Du, Jialiang / Lan, Zhiling / Liu, Yueyue / Liu, Yan / Li, Yanchao / Li, Xiangming / Guo, Tai

    Virus research

    2017  Volume 228, Page(s) 134–140

    Abstract: Mouse is one of the infection animal models for rotavirus. Since the optimal age of mouse sensitive to rotavirus infection thus far has not been unified, we elucidated clinical symptoms, immune responses and pathological changes of mice in different ages ...

    Abstract Mouse is one of the infection animal models for rotavirus. Since the optimal age of mouse sensitive to rotavirus infection thus far has not been unified, we elucidated clinical symptoms, immune responses and pathological changes of mice in different ages after challenged by murine rotavirus wild strain EDIM (Epidemic Diarrhea of Infant Mice) to provide data for the estimation. One-week-old, two-week-old, and three-week-old BALB/c mice were inoculated with EDIM in the challenge dose of 235 ID50, 470 ID50 and 705 ID50 respectively and were compared to mock-infected controls. Diarrhea illness, mobility, bodyweight were recorded, viral shedding and immune responses including serum IgA, fecal sIgA were detected, and small intestine tissue was evaluated for virus distribution and pathological changes. All the mice in one-week-old and two-week-old groups were completely unavoidable to be infected by EDIM and have been found to be malaise, activity reduced and even diarrhea, while three-week-old mice partly resist the challenge with 40% mice free from diarrhea. Meanwhile, EDIM infection has greater impact to the bodyweight of two-week-old group than those of one-week-old, three-week-old (0.9860 vs 1.2340, 1.2375g/day). One peak of virus shedding in three groups was observed in day 1-2 post infection, but the duration shortened with age increase. Feces sIgA in both two-week-old and three-week-old groups began to increase in day 4, 2-3days earlier than that in one-week-old group, and grow to the peak in day 8, which is about 2 fold of that in one-week-old group. Stronger serum IgA response was found in two-week-old group, it increased to the peak in day 15 and the level was 2 fold of three-week-old group and 4 fold of one-week-old group. The pathological changes included vacuolar degeneration, edema and congestion of intestinal wall, integrity destruction of enteric epithelium, and the changes relieved with the increase of age. Besides, rotavirus particles were found in small intestine tissues, especially in the surface and crypt of villi. In conclusion, the two-week-old mice were more sensitive to EDIM infection and initiated more effective immune response. In combination with that 14days old mice equals to 2 months infant when the first dose of rotavirus vaccine should be administrated, two-week-old mice is preferred to be used as infection model for the study of pathogenicity and immunogenicity of rotavirus.
    Language English
    Publishing date 2017-01-15
    Publishing country Netherlands
    Document type Journal Article
    ZDB-ID 605780-9
    ISSN 1872-7492 ; 0168-1702
    ISSN (online) 1872-7492
    ISSN 0168-1702
    DOI 10.1016/j.virusres.2016.12.001
    Database MEDical Literature Analysis and Retrieval System OnLINE

    More links

    Kategorien

  4. Article ; Online: Evaluation of oral Lanzhou lamb rotavirus vaccine via passive transfusion with CD4(+)/CD8(+) T lymphocytes.

    Du, Jialiang / Lan, Zhiling / Liu, Yueyue / Liu, Yan / Yu, Qingchuan / Li, Yanchao / Guo, Tai

    Virus research

    2016  Volume 217, Page(s) 101–106

    Abstract: Lanzhou Lamb derived Rotavirus (RV) Vaccine (namely LLR) for children is only used in China. Since there were no reports on evaluation of LLR, even the data of phase IV clinical trial, we proceed the evaluation of LLR through focusing on T-cell to ... ...

    Abstract Lanzhou Lamb derived Rotavirus (RV) Vaccine (namely LLR) for children is only used in China. Since there were no reports on evaluation of LLR, even the data of phase IV clinical trial, we proceed the evaluation of LLR through focusing on T-cell to investigate whether LLR could induce the potential function involving in protection as a vaccine. Four groups of nude mice were transfused with CD4(+)/CD8(+) T-cells isolated from LLR-immunized (primed) and LLR-unimmunized (naïve) mice via intraperitonea (i.p.) respectively. Consequently, the adoption mice were challenged with mice-origin wild rotavirus EDIM (Epizootic Diarrhea of Infant Mice) by intragastric administration. Series of fecal/serum samples were collected and viral shedding, then serum IgA/IgG and secreted IgA were assayed. Compared to the mice transfused with T lymphocytes from naïve mice, the nude mice transfused with CD4(+) T lymphocytes from primed mice induce fecal and serum IgA increasing more rapidly, and have a shorter duration of virus shedding too. Whereas, no significant difference in virus clearance was found between the mice transfused with CD8(+) T lymphocytes isolated from primed and naïve mice. Therefore, we cleared the distinct roles of transfused CD4(+)/CD8(+) T lymphocytes for rotavirus clearance in nude mice, that the viral clearance conducted by CD4(+) T lymphocytes. Meanwhile, it has ability to help induction of LLR specific immunogenicity. Comparing with the transfusion of cell from primed and naïve mice, LLR can induce CD4(+) T lymphocytes memory which is a potential index to reflect the immunogenicity and protection, while CD8(+) T lymphocytes remove rotavirus by CTL with little memory ability.
    Language English
    Publishing date 2016-06-02
    Publishing country Netherlands
    Document type Journal Article
    ZDB-ID 605780-9
    ISSN 1872-7492 ; 0168-1702
    ISSN (online) 1872-7492
    ISSN 0168-1702
    DOI 10.1016/j.virusres.2016.03.006
    Database MEDical Literature Analysis and Retrieval System OnLINE

    More links

    Kategorien

  5. Book ; Online: Scheduling Beyond CPUs for HPC

    Fan, Yuping / Lan, Zhiling / Rich, Paul / Allcock, William E. / Papka, Michael E. / Austin, Brian / Paul, David

    2020  

    Abstract: High performance computing (HPC) is undergoing significant changes. The emerging HPC applications comprise both compute- and data-intensive applications. To meet the intense I/O demand from emerging data-intensive applications, burst buffers are deployed ...

    Abstract High performance computing (HPC) is undergoing significant changes. The emerging HPC applications comprise both compute- and data-intensive applications. To meet the intense I/O demand from emerging data-intensive applications, burst buffers are deployed in production systems. Existing HPC schedulers are mainly CPU-centric. The extreme heterogeneity of hardware devices, combined with workload changes, forces the schedulers to consider multiple resources (e.g., burst buffers) beyond CPUs, in decision making. In this study, we present a multi-resource scheduling scheme named BBSched that schedules user jobs based on not only their CPU requirements, but also other schedulable resources such as burst buffer. BBSched formulates the scheduling problem into a multi-objective optimization (MOO) problem and rapidly solves the problem using a multi-objective genetic algorithm. The multiple solutions generated by BBSched enables system managers to explore potential tradeoffs among various resources, and therefore obtains better utilization of all the resources. The trace-driven simulations with real system workloads demonstrate that BBSched improves scheduling performance by up to 41% compared to existing methods, indicating that explicitly optimizing multiple resources beyond CPUs is essential for HPC scheduling.

    Comment: Accepted by HPDC 2019
    Keywords Computer Science - Distributed ; Parallel ; and Cluster Computing
    Subject code 000
    Publishing date 2020-12-09
    Publishing country us
    Document type Book ; Online
    Database BASE - Bielefeld Academic Search Engine (life sciences selection)

    More links

    Kategorien

To top