Data from: Enabling in situ visualization of large-scale cell simulations

Public

  • Studying cellular systems using computational models is essential in modern life sciences. These models can capture hundreds of millions of cells and generate petabytes of data, making visualization and analysis challenging. This paper presents an approach to enable in situ visualization and analysis of large-scale fluid-structure-interaction models on leadership-class systems. Our approach uses an in-line pipeline and in-transit method for real-time data interrogation and visualization during the main simulation. We demonstrate the feasibility of these approaches on a complex cell model with millions of components running on the Summit supercomputer and assess trade-offs in terms of run time, I/O contention, and data integrity. The proposed framework provides a valuable tool for both at-scale debugging and enabling scientific discovery, which would be difficult to achieve otherwise. We demonstrate data reduction at magnitudes up to 106 and set the groundwork for more efficient and effective explorative analysis of cellular mechanics. ... [Read More]

Total Size
97 files (1.42 MB)
Data Citation
  • Yousef, A., Rizzi, S., Insley, J., Mateevitsi, V., Randles, A. (2023). Data from: Enabling in situ visualization of large-scale cell simulations. Duke Research Data Repository. https://doi.org/10.7924/r49g5tm50
DOI
  • 10.7924/r49g5tm50
Publication Date
ARK
  • ark:/87924/r49g5tm50
Type
Format
Title
  • Data from: Enabling in situ visualization of large-scale cell simulations
This Dataset
Usage Stats