一个实用高性能PC集群的Linpack测试与分析*

HPL是Linpack Benchmark的并行版本,是高性能并行计算机系统性能评测的标准测试程序,对公正、全面、系统地评价机器的浮点性能具有重要意义。基于一个8节点的PC集群,测试了该集群的Linpack性能。根据实验数据,得出了集群性能与集群节点数之间近似线性关系的结论,并给出了处理器维数、问题规模与系统性能之间的关系。     

Fabricating nanostructures through a combination of nano-oxidation and wet etching on silicon wafers with different surface conditions

This study investigates the surface conditions of silicon wafers with native oxide layers (NOL) or hydrogen passivated layers (HPL) and how they influence the processes of nano-oxidation and wet etching. We also explore the combination of nano-oxidation and wet etching processes to produce nanostructures. Experimental results reveal that the surface conditions of silicon wafers have a considerable impact on the results of nano-oxidation when combined with wet etching. The height and width of oxides on NOL samples exceeded the dimensions of oxides on HPL samples, and this difference became increasingly evident with an increase in applied bias voltage. The height of oxidized nanolines on the HPL sample increased after wet etching; however, the width of the lines increased only marginally. After wet etching, the height and width of oxides on the NOL were more than two times greater than those on the HPL. Increasing the applied bias voltage during nano-oxidation on NOL samples increased both the height and width of the oxides. After wet etching however, the increase in bias voltage appeared to have little effect on the height of oxidized nanolines, but the width of oxidized lines increased. This study also discovered that the use of higher applied bias voltages on NOL samples followed by wet etching results in nanostructures with a section profile closely resembling a curved surface. The use of this technique enabled researchers to create molds in the shape of a silicon nanolens array and an elegantly shaped nanoscale complex structures mold.     

Atom transfer radical polymerization (ATRP): A versatile and forceful tool for functional membranes

The progress in atom transfer radical polymerization (ATRP) provides an effective means for the design and preparation of functional membranes. Polymeric membranes with different macromolecular architectures applied in fuel cells, including block and graft copolymers are conveniently prepared via ATRP. Moreover, ATRP has also been widely used to introduce functionality onto the membrane surface to enhance its use in specific applications, such as antifouling, stimuli-responsive, adsorption function and pervaporation. In this review, the recent design and synthesis of advanced functional membranes via the ATRP technique are discussed in detail and their especial advantages are highlighted by selected examples extract the principles for preparation or modification of membranes using the ATRP methodology.     

基于GNSS完好性的ADS-B防欺骗

为有效解决广播式自动相关监视（ADS-B）系统易被人为恶意虚假欺骗这一紧迫问题,提出了一种将全球导航卫星系统（GNSS）完好性信息与ADS-B报文完好性特征属性参数进行一致性比对的自主式防欺骗方法。设计了基于GNSS完好性信息的ADS-B自主式防欺骗算法流程,并利用ADS-B接收机接收的真实报文和模拟产生的虚假报文,对整个防欺骗过程实施了仿真实验。通过对真假ADS-B混合报文的解析,获得时间、位置和导航不确定类别（NUC）的值,根据时间和位置信息利用接收机自主完好性监视（RAIM）和高级接收机自主完好性监视（ARAIM）的完好性增强算法,获得水平保护级（HPL）的预测值,再将该预测值与NUC值进行一致性比对,从而判别ADS-B报文的真实性。仿真结果表明,提出的防欺骗方法能有效防止不含完好性信息或完好性信息与飞机位置不关联的虚假ADS-B报文的恶意欺骗,对虚假报文的识别率不会低于99%。     

A study of antioxidant properties of some varieties of grapes (Vitis vinifera L.)

Grapes (Vitis vinifera L.) are a major fruit crop in the world. Grapes seem to confer health benefits due to their antioxidant activity. We have evaluated the antioxidant potential of 11 grapes varieties from India and nearby Asian countries. The assays employed involve different levels of antioxidant action like ferric reducing antioxidant power, radical scavenging by 1,1-diphenyl-2-picrylhydrazyl, ferrylmyoglobin/2,2'-azobis-3-ethylbenzthiazoline-6-sulfonic acid, oxygen radical absorbance capacity (ORAC), and inhibition of lipid peroxidation. The total phenolic and flavonoids contents were also estimated. Our study indicates that cv. Mango is the most potent followed by Sharad Seedless. Ethanolic extracts were found to be more effective than aqueous extracts. Cv. Sharad Seedless, Mango, and Manikchaman also had high ORAC values. Their HPLC analysis showed the presence of various antioxidant polyphenols. In conclusion our studies identified some varieties of grapes with high antioxidant activities and showed that their high antioxidant potential may be due to their phenolic and flavonoid contents.     

国产异构系统上HPL的优化与分析

随着异构系统成为建造超级计算机的重要选择,如何让CPU与加速器协调工作以充分发挥异构系统的计算性能具有重要意义.HPL是高性能计算领域最重要的基准测试程序,传统面向纯CPU系统的HPL算法通过加速器加速矩阵乘法的做法已经无法取得很好的性能.针对这一问题,提出了基于国产处理器-国产加速器异构系统的HPL性能模型和多线程细...     

Human Platelet Lysate for Good Manufacturing Practice-Compliant Cell Production

Numerous cell-based therapeutics are currently being tested in clinical trials. Human platelet lysate (HPL) is a valuable alternative to fetal bovine serum as a cell culture medium supplement for a variety of different cell types. HPL as a raw material permits animal serum-free cell propagation with highly efficient stimulation of cell proliferation, enabling humanized manufacturing of cell therapeutics within a reasonable timeframe. Providers of HPL have to consider dedicated quality issues regarding identity, purity, potency, traceability and safety. Release criteria have to be defined, characterizing the suitability of HPL batches for the support of a specific cell culture. Fresh or expired platelet concentrates from healthy blood donors are the starting material for HPL preparation, according to regulatory requirements. Pooling of individual platelet lysate units into one HPL batch can balance donor variation with regard to essential platelet-derived growth factors and cytokines. The increasingly applied pathogen reduction technologies will further increase HPL safety. In this review article, aspects and regulatory requirements of whole blood donation and details of human platelet lysate manufacturing are presented. International guidelines for raw materials are discussed, and defined quality controls, as well as release criteria for safe and GMP-compliant HPL production, are summarized.     

面向国产异构系统的HPL异构协同设计

HPL是高性能计算广泛采用的Linpack测试软件包,传统HPL算法中,求解矩阵将以块为单位循环分布到所有处理器,由于国产加速器(China Accelerator)的底层矩阵乘接口仅支持定制接口,传统HPL算法已不适合CPU+China Accelerator异构系统,因此,必须基于定制接口完成矩阵分布细致划分与封装dPEM,以提供一个通用的HPL测试配置环境;同时,为了充分发挥国产异构系统的效率,设计了异构协同矩阵乘调度算法OA4MM,以提高国产异构系统的效率。实验验证了dPEM的有效性和OA4MM算法的高效性,OA4MM较传统的异构HPL调度算法性能提升近10%。     

Multi-communication layered HPL model and its application to GPU clusters

High-performance Linpack (HPL) is among the most popular benchmarks for evaluating the capabilities of computing systems and has been used as a standard to compare the performance of computing systems since the early 1980s. In the initial system-design stage, it is critical to estimate the capabilities of a system quickly and accurately. However, the original HPL mathematical model based on a single core and single communication layer yields varying accuracy for modern processors and accelerators comprising large numbers of cores. To reduce the performance-estimation gap between the HPL model and an actual system, we propose a mathematical model for multi-communication layered HPL. The effectiveness of the proposed model is evaluated by applying it to a GPU cluster and well-known systems. The results reveal performance differences of 1.1% on a single GPU. The GPU cluster and well-known large system show 5.5% and 4.1% differences on average, respectively. Compared to the original HPL model, the proposed multi-communication layered HPL model provides performance estimates within a few seconds and a smaller error range from the processor/accelerator level to the large system level.     

Complex version of high performance computing LINPACK benchmark (HPL)

This paper describes our effort to enhance the performance of the AORSA fusion energy simulation program through the use of high‐performance LINPACK (HPL) benchmark, commonly used in ranking the top 500 supercomputers. The algorithm used by HPL, enhanced by a set of tuning options, is more effective than that found in the ScaLAPACK library. Retrofitting these algorithms, such as look‐ahead processing of pivot elements, into ScaLAPACK is considered as a major undertaking. Moreover, HPL is configured as a benchmark, but only for real‐valued coefficients. We therefore developed software to convert HPL for use within an application program that generates complex coefficient linear systems. Although HPL is not normally perceived as a part of an application, our results show that the modified HPL software brings a significant increase in the performance of the solver when simulating the highest resolution experiments thus far configured, achieving 87.5 TFLOPS on over 20 000 processors on the Cray XT4. Copyright © 2009 John Wiley & Sons, Ltd.