Workshop III: HPC for Computationally and Data-Intensive Problems - IPAM

Workshop III: HPC for Computationally and Data-Intensive Problems

Part of the Long Program Science at Extreme Scales: Where Big Data Meets Large-Scale Computing
November 5 - 9, 2018

Schedule

All times in this Schedule are Pacific Time (PT)

Monday, November 5, 2018

Morning Session

08:00-08:55
Check-In/Light Breakfast (Hosted by IPAM)
08:55-09:00
Welcome and Opening Remarks
09:00-09:50
Nadia Heninger (University of Pennsylvania)
Computational challenges in applied cryptanalysis
10:00-10:15
Break
10:15-11:05
Torsten Hoefler (ETH Zurich)
High-Performance Communication for Machine Learning
11:15-11:30
Break
11:30-12:20
Wendy Cho (University of Illinois at Urbana-Champaign)
Massively Parallel Evolutionary Computation for Empowering Electoral Reform: Quantifying Gerrymandering via Multi-objective Optimization and Statistical Analysis

Afternoon Session

12:30-14:30
Lunch (on your own)
14:30-15:20
Rajat Monga (Google Inc.)
Challenges in Machine Learning Systems
15:30-16:00
Break
16:00-16:50
Alex Pothen (Purdue University)
A New Paradigm for Parallel Algorithm Design: Approximation
17:00-18:30
Poster Session & Reception (Hosted by IPAM)

Tuesday, November 6, 2018

Morning Session

08:00-09:00
Check-In/Breakfast (Hosted by IPAM)
09:00-09:50
Luke Olson (University of Illinois at Urbana-Champaign)
Pushing The Scalability Limits in Sparse Matrix Operations
10:00-10:15
Break
10:15-11:05
Vipin Kumar (University of Minnesota, Twin Cities)
Physics Guided Machine Learning: A New Paradigm for Modeling Dynamical Systems
11:15-11:30
Break
11:30-12:20
Madhav Marathe (University of Virginia)
Massively interacting bio-social systems: HPC-enable pervasive, personalized and precision analytics

Afternoon Session

12:30-14:30
Lunch (on your own)
14:30-15:20
Paul Morrin (University of Minnesota, Twin Cities)
The World Upside Down; NSF HPC produces the topography of the Earth
15:30-16:00
Break
16:00-16:50
Joachim Buhmann (ETH Zürich)
Robust algorithmics: a foundation for science?!

Wednesday, November 7, 2018

Morning Session

08:00-09:00
Check-In/Breakfast (Hosted by IPAM)
09:00-09:50
Karen Devine (Sandia National Laboratories)
Exploiting Scientific Software to Solve Problems in Data Analytics
10:00-10:15
Break
10:15-11:05
Ananth Grama (Purdue University)
Sampled Higher Order Distributed Non-Convex Optimization Techniques for Machine Learning
11:15-11:30
Break
11:30-12:20
Yann LeCun (New York University)
Computational Infrastructure Requirements for Deep Learning

Afternoon Session

12:30-14:30
Lunch (on your own)
14:30-15:20
Prabhat (Lawrence Berkeley Laboratory)
Data Analytics in the Exascale Era
15:30-16:00
Break
16:00-16:50
Allen Rush (AMD)
High Performance Mapping for Deep Learning Algorithms

Thursday, November 8, 2018

Morning Session

08:00-09:00
Check-In/Breakfast (Hosted by IPAM)
09:00-09:50
Tandy Warnow (University of Illinois at Urbana-Champaign)
Mathematical and Computational Challenges in Reconstructing Evolution
10:00-10:15
Break
10:15-11:05
Jennifer Listgarten (University of California, Berkeley (UC Berkeley))
Machine learning: from genetics to gene editing to protein optimization
11:15-11:30
Break
11:30-12:20
Petros Koumoutsakos (Harvard University)
High Performance Computing and Data science interfaces to predict, control and understand Fluid Mechanics

Afternoon Session

12:30-14:30
Lunch (on your own)
14:30-15:20
Siavash Mir arabbaygi (University of California, San Diego (UCSD))
Species tree inference and update on very large datasets using randomization, parallelization, and vectorization
15:30-16:00
Break
Contributed Talks
16:00-16:15
Erin Molloy (University of Maryland)
Scaling species tree estimation methods to large datasets using NJMerge
16:15-16:30
Anwesha Das (North Carolina State University)
Aarohi: Making Real-time Node Failure Prediction Feasible
16:30-16:45
Julija Zavadlav (Swiss Federal Institute of Technology of Zurich)
Data driven modeling in molecular dynamics
16:45-17:00
Torsten Hoefler (ETH Zurich)
Twelve ways to fool the masses when reporting performance of deep learning workloads
Printable Version