Ubiquitous Interactive Visualization of Large-Scale Simulations in Geosciences Over a Java-based Web-Portal Jonathan C. McLane (1), W. Walter Czech(2), David A. Yuen (1), Mike R.Knox (1), (3), S.Mark Wang(1), Jim B.S.G. Greensky(3), Erik O.D.Sevre (1). (1.)Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, MN 55455, U.S.A. (2.)Computer Science Institute, AGH University, Krakow, Poland (3.) Laboratory of Computational Science and Engineering, University of Minnesota, Minneapolis, MN, 55455, U.S.A. submitted to Concurrency: Practice and Experience, edited by G.C. Fox Abstract We have designed a web-based software system for real-time interactive visualization of results taken directly from large-scale simulations of 3-D mantle convection and other large- scale modeling efforts. This approach allows for intensive visualization sessions for a couple of hours as opposed to storing massive amounts of data in a storage system. Our data sets consist of 3-D data with over 10 million unknowns at each time-step to be used for volume rendering. Large scale interactive visualization on a display wall holding around 15 million pixels has already been accomplished. With the advent of the age of mobile web, we have now extended this tactic to hand-held devices, such as the OQO and Nokia N800, recently net-book computers, the iPHONE and the Android-powered system. We are developing web-based software in Java to extend the ubiquitous use of this system across long distances. The software is aimed at creating an interactive and functional application capable of running on multiple browsers by taking advantage of two AJAX-enabled web frameworks: Echo2 (http://echo.nextapp.com/site/echo2) and Google Web Toolkit (http://code.google.com/webtoolkit/). The software runs in two modes allowing for a user to control an interactive session or observe a session controlled by another user. Modular building of the system allows for components to be swapped out for new components so that other forms of visualization could be accommodated, such as molecular dynamics in mineral physics or 2-D data sets from regional convection models. 1. Introduction The relentless drive in pushing larger and faster massively parallel computers into the market , coupled with the introduction of GPUs , are stressing the dire need to visualize and analyze the enormous amount of numerical data being generated. The recent announcement of Blue Waters IBM system by the supercomputer center NCSA at Illinois has presaged the dawning of the age of petascale computing in just 3 years' time. The coming challenges are indeed quite daunting with the petascale machine Blue Water's RAM having around 1 petabytes and disk space over 20 Petabytes. We can easily expect to see the tsunami of output coming from solving partial differential equations, which may range between one billion and even one trillion grid points. With this gargantuan amount of numbers one cannot hope to proceed in the traditional way of post-processing. There is definitely a need for a new strategy for visualizing such large data sets or else we will be drowned under the torrential data being generated and not