Shift in thinking to address the 21st century hunger gap

The present water policy debate is dominated by the 30 yr old mission to secure water supply and sanitation to all people. The water needed to produce a nutritionally acceptable diet for one person is however 70 times as large as the amount needed for domestic water supply. The food security dilemma is largest in arid climate regions, a situation constituting a formidable challenge. It is suggested that an additional 5 600 km³/yr of consumptive water use will be needed to produce an adequate amount of food by 2050 – i.e almost a doubling of today’s consumptive use of 6800 km³/yr. Past misinterpretations and conceptual deficiencies show the importance of a shift in thinking. Combining the scale of the challenge and the time scale of the efforts to feed humanity and eradicate hunger leads to an impression of great urgency. This urgency strengthens the call for international research both for supporting agricultural upgrading, and for much better handling of issues of environmental sustainability. What stands out is the need of a new generation of water professionals, able to handle complexity and able to incorporate water implications of land use and of ecosystem health in integrated water resources management. It will for those reasons be essential and urgent to upgrade the educational system to producing this new generation.     

Linear Algebra Considerations for the Multi-Threaded Simulation of Mechanical Systems

A solution method suitable for the multi-threaded simulation ofmechanical systems represented in Cartesian coordinates isproposed and analyzed. In a state-space framework for thesolution of the Differential Algebraic Equations (DAE) ofMultibody Dynamics, the position/velocity stabilization and theacceleration computation are based on iterative solvers applied toequivalent reduced problems. The most in-depth computationalaspect analyzed is the preconditioning, i.e., the direct solutionof the reduced systems. Provided a topology index reduction is first applied to the model, the effort for the direct solution of the reduced systems is shown to be of order O(N _J ), where N _J is the number of joints in the model. The recurring theme of thepaper is the central role that the topology of the mechanicalsystem plays in the overall performance of the numericalsimulation. Based on the topology of the model, parallelcomputational threads can be established to start in the equationformulation and continue through the iterative numericalalgorithms employed for the numerical solution. Task schedulingthese parallel threads is expected to redeem real-time performancefor certain classes of complex applications.     

一种应用于二次布局的有效划分方法

提出了一种基于二次布局的结合MFFC结群和hMETIS划分的算法.实验表明:这种方法能得到很好的布局结果,但是运行消耗的时间比较长.为了缩短划分在二次布局中运行的时间,提出了一种改进的结群算法IMFFC,用它在二次布局中做划分.与前者相比较,这种方法虽然布局质量稍差,但速度更快.     

COPART:一种面向约束条件的自适应软硬件划分算法

本文介绍一种将自适应算法、遗传算法和列表调度算法结合起来应用于软硬件划分问题的算法COPARTART，并提出了一种基于约束条件的开销系数自适应调整方法，该方法所获得的划分结果能够良好体现设计目标。     

周期性业务流通过ATM复用器的时延抖动分析

ＡＴＭ通信网中需要传送多种业务，其中有一类是具有连续比特流的业务，这种业务的发送信元到达时间是周期性的，本文分析了一种周期性信元流通过ＡＴＭ复用器的排队模型，得出信元流排队时延的概率分布，时延抖动的自相关函数的功率谱。     

基于集群计算机的保护定值在线校核

采用分布式计算技术，研究并实现了基于集群计算机的保护定值在线校核并行计算方法。首先分析了保护定值校核计算的计算量和计算特点，接着对可能采用的并行计算方法进行比较，最后选择了基于电网分块的并行计算方法。采用协调机加计算子机的计算模式，在协调机上执行分块算法，通过基于主导节点的图划分方法把电网分成几块，每台计算子机只校核1块电网内的保护定值，最后由协调机收集汇总计算结果。现场应用证明，该并行计算方法正确、有效。     

Experiences with a parallel algorithm for data flow analysis

We have designed a family of parallel data flow analysis algorithms for execution on distributed-memory MIMD machines, based on general-purpose, hybrid algorithms for data flow analysis [Marlowe and Ryder 1990]. We exploit a natural partitioning of the hybrid algorithms and explore a static mapping, dynamic scheduling strategy. Alternative mapping-scheduling choices and refinements of the flow graph condensation used are discussed. Our parallel hybrid algorithm family is illustrated on Reaching Definitions, although parallel algorithms also exist for many interprocedural (e.g., Aliasing) and intraprocedural (e.g., Available Expressions) problems [Marlowe 1989]. We have implemented the parallel hybrid algorithm for Reaching Definitions on an Intel iPSC/2. Our empirical results suggest the practicality of parallel hybrid algorithms.An earlier version of this paper was presented at Supercomputing '90.The research reported here was supported, in part, by the New Jersey Commission on Science and Technology and the CAIP Center's Industrial Members, by Siemens Research Corporation and by National Science Foundation grant CCR-8920078.     

Effect of partitioning on microstructure and properties of a medium carbon Ti Mo bearing high strength Q&P steel

In order to investigate the effect of partitioning procedures on microstructure and properties of a medium carbon Ti Mo bearing steel, the salt bath experiment, field emission SEM, TEM and tensile tests as well as wear tests were utilized to clarify the microstructural evolution and property changes treated by different partitioning parameters. The results show that the microstructures consist of tempered lath martensite, cementite and (Ti, Mo)C particles. With prolonging partitioning time or increasing partitioning temperature, the amount of laths martensite decreases and thicker martensite plates with blunt boundaries appear. In addition, with increasing partitioning temperature from 310℃ to 400℃, the tensile strength, hardness and low temperature toughness are simultaneously decreased, and the reduction reaches up to about 300MPa, 100HV and 20J, respectively. Finally, the wear performance of the samples partitioned below Ms is obviously better than that of samples partitioned above Ms. The morphology of worn surface of samples partitioned below Ms is characterized by plastic fatigue morphology, while the worn surfaces of samples partitioned above Ms are mainly dominated by grooves.     

Data decomposition using independent component analysis with prior constraints

In many data analysis problems, it is useful to consider the data as generated from a set of unknown (latent) generators or sources. The observations we make of a system are then taken to be related to these sources through some unknown function. Furthermore, the (unknown) number of underlying latent sources may be less than the number of observations. Recent developments in independent component analysis (ICA) have shown that, in the case where the unknown function linking sources to observations is linear, such data decomposition may be achieved in a mathematically elegant manner. In this paper, we extend the general ICA paradigm to include a very flexible source model, prior constraints and conditioning on sets of intermediate variables so that ICA forms one part of a hierarchical system. We show that such an approach allows for efficient discovery of hidden representation in data and for unsupervised data partitioning.