ISAV 2015: In Situ Infrastructures for Enabling Extreme-scale Analysis and Visualization

Held in conjunction with
SC15: The International Conference on High Performance Computing, Networking, Storage, and Analysis.

Monday, 16 November 2015, 2:00 - 5:30p, Hilton Salon D, 500 East 4th Street, Austin, TX


ISAV 2015 Proceedings at ACM: https://dl.acm.org/citation.cfm?id=2828612

Workshop Theme

This workshop brings together researchers, developers and practitioners from industry, academia, and government laboratories who use in situ methods in extreme-scale, high performance computing. The goal is to present existing in-situ infrastructures, reference examples in a range of science and engineering applications, to discuss topics like opportunities presented by new architectures; existing infrastructure needs, requirements, and gaps; and experiences to foster and enable in situ analysis and visualization.

The considerable interest in the HPC community regarding in situ analysis and visualization is due to several factors. First is an I/O cost savings, where data is analyzed/visualized while being generated, without first storing to a filesystem. Second is the potential for increased accuracy, where fine temporal sampling of transient analysis might expose some complex behavior missed in coarse temporal sampling. Third is the ability to use all available resources, CPU's and accelerators, in the computation of analysis products.

In situ processing is still a relatively new idea, and up until recently, most implementations have been ad hoc, proof-of-concept prototypes. However, several in situ infrastructure implementations have emerged. ParaView and VisIt both provide tools for in situ analysis and visualization. ParaView Catalyst can be linked to a simulation, allowing the simulation to share data with Catalyst for visualization. Similar capabilities are available within VisIt with the libsim library. Both Catalyst (through Live) and libsim enable the opposite flow of information, sending data from the client to the simulation, enabling the possibility of simulation steering. ADIOS and GLEAN allow simulations to adopt in situ techniques by leveraging their advanced I/O infrastructures that enable co-analysis pipelines rather than changing the simulator. The non-intrusive integration provides resilience to third party library bugs and possible jitter in the simulation.

Program

2:00 - 2:10pISAV Opening & Welcome (Bethel, Vishwanath, Weber, Wolf)
2:10 - 2:45p Keynote: Challenges in the In Situ Methods/Infrastructure in the Extreme-Scale Fusion Code XGC1 (C.-S. Chang)
2:45 - 3:00p Lightning Talks (4 minute talks & 3 minutes discussion; Session chair/moderator: Wolf)
  • Tightly-coupling In Situ Terms With Concepts While They Are Still In Transit (Childs)
  • Exploring a Unified Task-Based Approach for In Situ Analysis and Visualization (Sweeney, McCormick, Sohl & Tong)
  • In Situ Analysis and Visualization for Rotor Aeromechanics Simulations (Jimenez, Singh & Bauer)
3:00 - 3:30pCoffee Break
3:30 - 4:25pPerspectives on In Situ Coupling Frameworks (12 minute talks followed by 7 minutes questions/discussion; session chair: Vishwanath)
  • Loosely Coupled In Situ Visualization: A Perspective on Why It's Here to Stay (Kress et al.)
  • An Approach to Lowering the In Situ Visualization Barrier (Fogal & Krüger)
  • In Situ Analysis as a Parallel I/O Problem (Ziegeler et al.)
  • Lessons Learned from Building In Situ Coupling Frameworks (Dorier et al.)
  • Discussion
4:25 - 5:20pApplications of In Situ Analysis (12 minute talks followed by 7 minutes questions/discussion; session chair: Weber)
  • ParaView Catalyst: Enabling In Situ Data Analysis and Visualization (Ayachit et al.)
  • Strawman: A Batch In Situ Visualization and Analysis Infrastructure for Multi-Physics Simulation Codes (Larsen et al.)
  • Infrastructure for In Situ System Monitoring and Application Data Analysis (Brandt et al.)
  • Enabling Adaptive Scientific Workflows Via Trigger Detection (Salloum et al.)
  • Discussion
5:20-5:30pWorkshop Feedback from Participants and Closing Statement (Bethel)

Proceedings

The proceedings are published in the ACM Digital Library.

Participation/Call for Papers

We invite short (4-page) papers that identify opportunities, challenges and case studies/best practices for in situ analysis and visualization. These papers could propose actions, or provide position, or experience reports on in situ analysis and visualization.

Areas of interest for ISAV, include, but are not limited to:

In situ infrastructures
  • Current Systems: production quality, research prototypes
  • Opportunities
  • Gaps
System resources, hardware, and emerging architectures
  • Enabling Hardware
  • Hardware and architectures that provide opportunities for In situ processing, such as burst buffers, staging computations on I/O nodes, sharing cores within a node for both simulation and in situ processing
Examples/Case studies
  • Best practices
  • Analysis: feature detection, statistical methods, temporal methods, geometric methods
  • Visualization: information visualization, scientific visualization, time-varying methods
  • Data reduction/compression
Simulation
  • Integration:data modeling, software-engineering
  • Resilience: error detection, fault recovery
  • Workflows for supporting complex in situ processing pipelines
Requirements
  • Preserve important elements
  • Significantly reduce the data size
  • Flexibility for post-processing exploration

Submitting Papers

Submissions are limited to 4 pages in the ACM format (see here for templates). The 4-page limit includes figures, tables, and appendices, but does not include references, for which you may use up to one additional page. Please submit your paper at the ISAV 2015 EasyChair submission page.

Peer review process

All submissions will undergo a peer-review process consisting of three reviews by experts in the field.

Publication in proceedings, presentation at the workshop

All submissions that receive favorable reviews will be included as part of the workshop proceedings, which will be published along with other SC15 workshop proceedings in the ACM Digital Library and IEEE Xplore.

Subject to the constraints of workshop length, some subset of the accepted publications will be invited to give a brief oral presentation at the workshop. The exact number of such presentations and their length will be determined after the review process has been completed.

Timeline/Important Dates

1 Aug 201516 Aug 2005 Paper submission deadline
1 Sep 201516 Sep 2015 Author notification
15 Sep 201530 Sep 2015 Camera ready copy due
mid-October 2015 Final program posted to ISAV web page
16 Nov 2015 ISAV workshop at SC15

Committees

Organizing Committee
E. Wes Bethel, Lawrence Berkeley National Laboratory, USA
Venkatram Vishwanath, Argonne National Laboratory, USA
Gunther H. Weber, Lawrence Berkeley National Laboratory, USA
Matthew Wolf, Georgia Institute of Technology, USA
Program Committee
Utkarsh Ayachit, Kitware Inc., USA
Earl P.N. Duque, Intelligent Light, USA
Nicola Ferrier, Argonne National Laboratory, USA
Burlen Loring, Lawrence Berkeley National Laboratory, USA
Dmitriy Morozov, Lawrence Berkeley National Laboratory, USA
Patrick O'Leary, Kitware Inc., USA
Manish Parashar, Rutgers, USA
Alex Sim, Lawrence Berkeley National Laboratory, USA
Brad Whitlock, Intelligent Light, USA
Kesheng (John) Wu, Lawrence Berkeley National Laboratory, USA
Early Career Program Committee
Jai Dayal, Georgia Institute of Technology, USA
Anshuman Goswami, Georgia Institute of Technology, USA
James Kress, University of Oregon, USA
Matt Larsen, University of Oregon, USA
Preeti Malakar, Argonne National Laboratory, USA
Khairi Reda, Argonne National Laboratory, USA
Silvio Rizzi, Argonne National Laboratory, USA

Contact

For questions about the workshop, please contact E. Wes Bethel at ewbethel at lbl dot gov