Mapping Parallel Application Communication Topology to Rhombic Overlapping-Cluster Multiprocessors

This paper extends research into rhombic overlapping-connectivity interconnection networks into the area of parallel applications. As a foundation for a shared-memory non-uniform access bus-based multiprocessor, these interconnection networks create overlapping groups of processors, buses, and memories, forming a clustered computer architecture where the clusters overlap. This overlapping-membership characteristic is shown to be useful for matching parallel application communication topology to the architecture's bandwidth characteristics. Many parallel applications can be mapped to the architecture topology so that most or all communication is localized within an overlapping cluster, at the low latency of processor direct to cache (or memory) over a bus. The latency of communication between parallel threads does not degrade parallel performance or limit the graininess of applications. Parallel applications can execute with good speedup and scaling on a proposed architecture which is designed to obtain maximum advantage from the overlapping-cluster characteristic, and also allows dynamic workload migration without moving the instructions or data. Scalability limitations of bus-based shared-memory multiprocessors are overcome by judicious workload allocation schemes, that take advantage of the overlapping-cluster memberships. Bus-based rhombic shared-memory multiprocessors are examined in terms of parallel speedup models to explain their advantages and justify their use as a foundation for the proposed computer architecture. Interconnection bandwidth is maximized with bi-directional circular and segmented overlapping buses. Strategies for mapping parallel application communication topologies to rhombic architectures are developed. Analytical models of enhanced rhombic multiprocessor performance are developed with a unique bandwidth modeling technique, and are compared with the results of simulation.     

Query-driven visualization of time-varying adaptive mesh refinement data

The visualization and analysis of AMR-based simulations is integral to the process of obtaining new insight in scientific research. We present a new method for performing query-driven visualization and analysis on AMR data, with specific emphasis on time-varying AMR data. Our work introduces a new method that directly addresses the dynamic spatial and temporal properties of AMR grids that challenge many existing visualization techniques. Further, we present the first implementation of query-driven visualization on the GPU that uses a GPU-based indexing structure to both answer queries and efficiently utilize GPU memory. We apply our method to two different science domains to demonstrate its broad applicability.     

Detection of environmental changes from network structure

We studied the networks of the temperature record in the atmosphere. They are made by the strength of the synchronization, including the delay between temperatures at locations on Earth. We consider these locations as nodes, and we consider a pair of locations as a link if the synchronization between them is stronger than a threshold. The network is scale-free, which is thought to contribute to the stability of the climate. This work was presented in part at the 12th International Symposium on Artificial Life and Robotics, Oita, Japan, January 25–27, 2007     

Leveraging online communities in fostering adaptive schools

There has long been a call for schools to prepare students for the twenty-first century where skills and dispositions differ significantly from much of what has historically characterized formal education. The knowledge based economy calls for policy and pedagogical efforts that would transform schools. Schools are to foster communities of learners. This paper suggests that para-communities may be points of leverage in the fostering of adaptive schools. A critical analysis is done on the differences between para-communities (such as online communities) and schools; and an argument is made that they each serve differing goals and should be left distinct because they achieve different societal and economic demands.     

Optimization strategy in the development process with application to cast components

At the development of commercial vehicles, it is necessary to develop cost-optimized components with respect to their entire lifetime. This requires a product development process that utilizes the possibility to simulate the product performance using modeling of different degrees of sophistication at any predefined level-of-detail (LOD). For example, the mechanical properties of a cast component depend strongly on the casting process, the simulation of which requires complex modeling at a high LOD. In this article, we propose an optimization strategy that is based on the LOD concept to develop cost-effective components. A sizing optimization of a cast model component is carried out with respect to both the mechanical and thermal response to exemplify the process. In particular, a (novel) global measure of the temperature in energy norm is introduced as an indicator of the “production fitness” to ensure high quality of the material properties. This measure is the thermal counterpart of the well-accepted mechanical strain energy.     

Stroboscopic image capture: reducing the dose per frame by a factor of 30 does not prevent beam-induced specimen movement in paraffin

Beam-induced specimen movement may be the major factor that limits the quality of high-resolution images of organic specimens. One of the possible measures to improve the situation that was proposed by Henderson and Glaeser [Ultramicroscopy 16 (1985) 139-150], which we refer to here as "stroboscopic image capture", is to divide the normal exposure into many successive frames, thus reducing the amount of electron exposure--and possibly the amount of beam-induced movement--per frame. The frames would then be aligned and summed. We have performed preliminary experiments on stroboscopic imaging using a 200-kV electron microscope that was equipped with a high dynamic range Charge-coupled device (CCD) camera for image recording and a liquid N2-cooled cryoholder. Single-layer paraffin crystals on carbon film were used as a test specimen. The ratio F(g)/F(0) of paraffin reflections, calculated from the images, serves as our criterion for the image quality. In the series that were evaluated, no significant improvement of the F(image)(g)/F(image)(0) ratio was found, even though the electron exposure per frame was reduced by a factor of 30. A frame-to-frame analysis of image distortions showed that considerable beam-induced movement had still occurred during each frame. In addition, the paraffin crystal lattice was observed to move relative to the supporting carbon film, a fact that cannot be explained as being an electron-optical effect caused by specimen charging. We conclude that a significant further reduction of the dose per frame (than was possible with this CCD detector) will be needed in order to test whether the frame-to-frame changes ultimately become small enough for stroboscopic image capture to show its potential.     

Time dependent processing in a parallel pipeline architecture

Pipeline architectures provide a versatile and efficient mechanism for constructing visualizations, and they have been implemented in numerous libraries and applications over the past two decades. In addition to allowing developers and users to freely combine algorithms, visualization pipelines have proven to work well when streaming data and scale well on parallel distributed-memory computers. However, current pipeline visualization frameworks have a critical flaw: they are unable to manage time varying data. As data flows through the pipeline, each algorithm has access to only a single snapshot in time of the data. This prevents the implementation of algorithms that do any temporal processing such as particle tracing; plotting over time; or interpolation, fitting, or smoothing of time series data. As data acquisition technology improves, as simulation time-integration techniques become more complex, and as simulations save less frequently and regularly, the ability to analyze the time-behavior of data becomes more important. This paper describes a modification to the traditional pipeline architecture that allows it to accommodate temporal algorithms. Furthermore, the architecture allows temporal algorithms to be used in conjunction with algorithms expecting a single time snapshot, thus simplifying software design and allowing adoption into existing pipeline frameworks. Our architecture also continues to work well in parallel distributed-memory environments. We demonstrate our architecture by modifying the popular VTK framework and exposing the functionality to the ParaView application. We use this framework to apply time-dependent algorithms on large data with a parallel cluster computer and thereby exercise a functionality that previously did not exist.     

Compositional pattern producing networks: A novel abstraction of development

Natural DNA can encode complexity on an enormous scale. Researchers are attempting to achieve the same representational efficiency in computers by implementing developmental encodings, i.e. encodings that map the genotype to the phenotype through a process of growth from a small starting point to a mature form. A major challenge in in this effort is to find the right level of abstraction of biological development to capture its essential properties without introducing unnecessary inefficiencies. In this paper, a novel abstraction of natural development, called Compositional Pattern Producing Networks (CPPNs), is proposed. Unlike currently accepted abstractions such as iterative rewrite systems and cellular growth simulations, CPPNs map to the phenotype without local interaction, that is, each individual component of the phenotype is determined independently of every other component. Results produced with CPPNs through interactive evolution of two-dimensional images show that such an encoding can nevertheless produce structural motifs often attributed to more conventional developmental abstractions, suggesting that local interaction may not be essential to the desirable properties of natural encoding in the way that is usually assumed.