Organizers

• Tom Peterka, "tpeterka" AT "mcs.anl.gov"
• Rob Ross, "rross" AT "mcs.anl.gov"
• Rusty Lusk, "lusk" AT "mcs.anl.gov"
• Pete Beckman, "beckman" AT "mcs.anl.gov"

Abstract

This workshop explores computer science topics at the intersection of storage, analysis, and data-intensive computational science. Traditionally, computation, storage, and analysis were viewed as three separate tasks, but as supercomputers and data increase in size and complexity, this isolated view is no longer practical. By considering these topics together, we hope to uncover potential efficiency and scalability, and see the connections between these areas as the week progresses.

Within this broad scope, relevant topics can include data models, run-time and postprocessing analysis and visualization techniques, I/O systems and optimizations, data mining and machine learning, and the application of all of the above in science codes. However, we welcome you to think beyond these topics and propose new problems and solutions. While we value solutions that are ready to use today, we encourage research that is 1-5 years out as well. The workshop features ample opportunity for discussion following each set of talks, as well as open time for one-on-one collaborations.

Agenda

Day 0 - Sunday, July 24

* 7:00-8:30 PM reception dinner

Day 1 - Monday, July 25

* Breakfast buffet 7:30-8:30

* Morning

Introduction (Tom Peterka, chair)
o 8:30-9:00 Opening and Introduction (Tom Peterka)
o 9:00-10:00 Research Introductions. All participants give short (5 min) talks on their research areas and data challenges they face.
o 10:00-10:15 Short break, discussions continue, refreshments served
o 10:15-11:45 Wrap up introductions

Data Models (Dries Kimpe, chair)
o 11:45-12:15 MOAB (Tim Tautges)

* Lunch buffet 12:15-1:15

* Afternoon

Data Models continued
o 1:30-2:00 The VTK Data Model (Berk Geveci)
o 2:00-2:30 Damsel: A Data Model Storage Library for Exascale Science (Saba Sehrish)
o 2:30-2:45 Short break, discussions continue, refreshments served
o 2:45-3:15 Parallel HDF5 (Quincey Koziol)
o 3:45-4:15 Discussion

* Evening

o 6:00-7:30 Working Buffet Dinner: discuss ideas raised by the day's presentations and discuss possibilities for Tuesday's hands-on work


Day 2 - Tuesday, July 26

* Breakfast buffet 7:30-8:30

* Morning

Data Analysis Infrastructures (Rob Ross, chair)
o 8:30-9:00 GLEAN (Venkat Vishwanath)
o 9:00-9:30 Large-Scale Flow Analysis with DStep (Wes Kendall)
o 9:30-10:00 Data Movement Support for Analysis (Tom Peterka)
o 10:00-10:15 Short break, discussions continue, refreshments served
o 10:15-10:45 Discussion

Computation in Science Codes (Tom Peterka, chair)
o 10:45-11:15 Cosmological Simulations: The HACC Framework (Salman Habib)
o 11:15-11:45 Data Structures in the CSU GCRM (Don Dazlich)
o 11:45-12:15 Combustion in S3D (Ray Grout)

* Lunch buffet 12:15-1:15

* Afternoon

Data Analysis in Science Codes (Saba Sehrish, chair)
o 1:30-2:00 Analysis of Cosmological Simulations (Katrin Heitmann)
o 2:00-2:30 Data Analysis and Visualization of Petascale Combustion Science Simulation Data (Janine Bennett)
o 2:30-2:45 Short break, discussions continue, refreshments served
o 2:45-3:15 Geospatiotemporal Data Mining for Large Earth Science Data Sets (Jitu Kumar)
o 3:15-3:45 Discussion
o 3:45-5:00 Hands-on Session I: one-on-one and group discussions, time is left open for data experts to meet one-on-one to discuss tools, techniques, and research

* Evening

o 6:00-7:30 Working Buffet Dinner: discuss ideas raised by the day's presentations and plan for late night hacking
o 7:30-Late Meeting space open for late-night hacking and collaborative work


Day 3 - Wednesday, July 27

* Breakfast buffet 7:30-8:30

* Morning

Analysis and Visualization Techniques (Attila Gyulassy, chair)
o 8:30-9:00 Quantifiable Raw Data Reduction for Post-Simulation Scientific Analysis (Jon Woodring)
o 9:00-9:30 Integrated Data Management Analysis and Visualization (Valerio Pascucci)
o 9:30-10:00 Understanding Climate Change: Opportunities and Challenges for Data Intensive Computational Science (Vipin Kumar)
o 10:00-10:15 Short break, discussions continue, refreshments served
o 10:15-10:45 FastBit Indexing for Searching and Analyzing Massive Data (John Wu)
o 10:45-11:15 Information Theory for Visualization and Analysis (Teng-Yok Lee)
o 11:15-11:45 Parallel Computation of Morse-Smale Complexes (Attila Gyulassy)
o 11:45-12:15 Discussion

* Lunch buffet 12:15-1:15

* Afternoon

o 1:30-5:00 Hands-on Session II: one-on-one and group discussions, time is left open for data experts to meet one-on-one to discuss tools, techniques, and research
o 2:30-2:45 Short break, discussions continue, refreshments served

* Evening

o 6:00-7:30 Working Buffet Dinner: discuss progress in hand-on sessions, issues, and plan for late-night hacking
o 7:30-Late Meeting space open for late-night hacking and collaborative work


Day 4 - Thursday, July 28

* Breakfast buffet 7:30-8:30

* Morning

Future Data Systems and Software (Tom Peterka, chair)
o 8:30-9:00 Extreme Scale System Center File System Research (Brad Settlemyer)
o 9:00-9:30 Exascale Systems Software (Rob Ross)
o 9:30-10:00 Discussion
o 10:00-10:15 Short break, discussions continue, refreshments served
o 10:15-10:45 Open Discussion. All attendees discuss the workshop, perceived current and future challenges for exascale computing.

* Lunch buffet 12:15-1:15

* Afternoon

o 1:30-3:00 Hands-on Session III: one-on-one and group discussions, time is left open for data experts to meet one-on-one to discuss tools, techniques, and research
o 2:30-2:45 Short break, discussions continue, refreshments served
o 3:00 Meeting adjourns

* Evening (for those who have not checked out yet)

o 6:00-7:30 Working Buffet Dinner: discuss future collaborations
o 7:30-Late Meeting space open for late-night hacking and collaborative work

Day 5 - Friday, July 29 (for those who have not checked out yet)

* 7:30-8:30 Breakfast buffet

* Boxed to-go lunches available for those who requested them earlier

Sponsors