Abstract

We propose a new visualization method for massive supercomputer simulations. The key idea is to scatter multiple omnidirectional cameras to record the simulation via in situ visualization. After the simulations are complete, researchers can interactively explore the data collection of the recorded videos by navigating along a path in four-dimensional spacetime. We demonstrate the feasibility of this method by applying it to three different fluid and magnetohydrodynamics simulations using up to 1,000 omnidirectional cameras.

Highlights

  • Each time the user specifies a new camera position or switches to a new visualization method, the 4D Street Viewer retrieves the corresponding file from the data collection

  • The MP4 conversion by ffmpeg from the PNG images is the only stage wherein lossy compression is applied in our procedure. (The mapping to an omnidirectional image from the six directional images described incorporates pixel interpolations, but the image deterioration by the interpolation is negligibly small)

  • The largest number of the omnidirectional cameras scattered in the test simulations was 1000, that are for the vortex ring simulation

Read more

Summary

INTRODUCTION

Visualization is indispensable to comprehend the time development of complex phenomena captured as numerical data by supercomputer simulations. When we place C cameras and apply V kinds of visualizations on each of them, we obtain C × V video files in total as a result of simulation (Fig. 2B). The periodic boundary condition is assumed for all (x, y and z) directions For this simulation, we have tried three different configurations of the omnidirectional cameras; (Cx, Cy, Cz) = (10, 10, 10), (20,10,5) and (3, 2, 1) with three visualization methods; (1) isosurface of the enstrophy density e = |∇ × v|2, where v is the flow velocity; (2) arrow glyphs of v; (3) stream lines of v. Each time the user specifies a new camera position or switches to a new visualization method, the 4D Street Viewer retrieves the corresponding file from the data collection.

DISCUSSION
CONCLUSIONS
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.