Time Dependent Processing in a Parallel Pipeline Architecture

Abstract

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.

Skeletons for distributed topological computation

“…While this consists of multiple time-varying scalar and vector fields, for present purposes we work with two fields, 2 The inquiry function passed as a parameter to the skeleton is an unfortunate implementation detail. We use Eden's 'remote data' capability to avoid the overhead of Eden's default hierarchical data transfers.…”

“…In many cases the data is part of a time series in which case time can be treated as a further dimension, allowing for example, the interpolation of models between discrete simulation steps [2]. In general we consider the phenomena of interest to be finite samples taken from the continuous function f : R m(+1) → R n , where m is the spatial domain dimension.…”

Fine-Grained Provenance of Users’ Interpretations in a Collaborative Visualization Architecture

“…Using metadata, users can select a region or multiple regions to process, for example this allowed Ahrens et al to visualize largescale datasets using parallel data streaming (Ahrens et al, 2001). In addition to regional information, a time dimension can be added to metadata, adding time control to the visualization pipeline (Biddiscombe et al, 2007). The usefulness of metadata was further developed with the introduction of query-driven visualization (Stockinger et al, 2005;Gosink et al, 2008).…”

Multiverse Data-Flow Control

“…The extended data-flow concept presented in this paper alleviates this apparent conflict. Biddiscombe et al [BGMT07] extend the VTK data-flow to support nodes which require multiple timesteps. Their approach does not perform caching when processing multiple timesteps, which greatly simplifies the required logic.…”