The current availability of thousands of processors at many high performance computing centers has made it feasible to carry out, in near real time, interactive visualization of 3D mantle convection temperature fields, using grid configurations having 10-100 million unknowns. We will describe the technical details involved in carrying out this endeavor, using the facilities available at the Laboratory of Computational Science and Engineering (LCSE) at the University of Minnesota. These technical details involve the modification of a parallel mantle convection program, ACuTEMan; the usage of client-server socket based programs to transfer upwards of a terabyte of time series scientific model data using a local network; a rendering system containing multiple nodes; a high resolution PowerWall display, and the interactive visualization software, DSCVR. We have found that working in an interactive visualizastion mode allows for fast and efficient analysis of mantle convection results.

The Dementia Rating Scale (DRS; Mattis, 1976, 1988) is commonly used in the assessment of dementia, although little is known about the relationship of performance on this test to specific cognitive deficits in Alzheimer's disease (AD). Additionally, cognitive profiles have not been investigated across different levels of dementia as determined by the DRS. A sample of 133 individuals diagnosed with possible or probable AD was administered the DRS as part of a comprehensive neuropsychological evaluation. Composite scores for the cognitive domains of attention, executive functioning, visuospatial skills, language abilities, immediate recall, and delayed memory were derived by averaging demographically corrected T scores of key measures. Individual domain scores were also averaged to develop a global index score. Pearson correlations between composite and total DRS scores were highly significant (p<.001) for all domains and the global index score, with the exception of delayed memory, which showed a floor effect. When the sample was divided into mild and moderate-to-severe groups to examine the effects of disease severity on the relationship between the DRS and standard neurocognitive domain scores, the resulting mean neuropsychological profile scores were significantly different while maintaining a parallel pattern of impairment across domains. Results demonstrate the relationship between the DRS and standard cognitive domain functions, which appears to underscore the validity and robustness of the DRS in characterizing patterns of cognitive impairment across the AD spectrum.

We have developed a new strategy and espouse a novel paradigm for large-scale computing and real-time interactive visualization. This philosophy calls for intense interactive sessions for a couple of hours at a time at the expense of storing data on many disk drives during regular or heroic runs on massively parallel systems. We have already carried out successfully real-time volume-rendering visualization by employing hundreds of processors for a grid with over 25 million unknowns. Both Cartesian and spherical 3D mantle convection are visualized. The volume-rendered images are viewed on a large display device, with many panels holding around 13 million pixels. We will employ a software strategy involving an hierarchical rendering service, which will have as software an Ajax interface for interactive visualization of large data sets on many different platforms from desktop PC's to hand-held devices, such as the OQO and the Nokia N-800. An option for stereo viewing is also implemented. We have installed a user interface as web application, using Java and Ajax framework in order to achieve over the Internet reasonable accessibility to our ongoing runs. Our goal is to expand the array of interactive devices, which will make it feasible to carry out ubiquitous visualization and monitoring of largescale simulations or onsite events and to allow for collaborations across oceans.

An extreme form of pipelining of the Piecewise-Parabolic Method (PPM) gas dynamics code has been used to dramatically increase its performance on the new generation of multicore CPUs. Exploiting this technique, together with a full integration of the several data post-processing and visualization utilities associated with this code has enabled numerical experiments in computational fluid dynamics to be performed interactively on a new, dedicated system in our lab, with immediate, user controlled visualization of the resulting flows on the PowerWall display. The code restructuring required to achieve the necessary CPU performance boost, as well as the parallel computing methods and systems used to enable interactive flow simulation are described. Requirements for these techniques to be applied to other codes are discussed, and our plans for tools that will assist programmers to exploit these techniques are briefly described. Examples showing the capability of the new system and software are given for applications in turbulence and stellar convection.