Restaurants name, place and time and workshop buses schedule.
This event is supported by INRIA, UIUC, NCSA, ANL, as well as by EDF
Main Topics | Schedule | Speaker | Affiliation | Type of presentation | Title (tentative) | Download |
|
|
|
|
|
|
|
Dinner Before the Workshop | 8:00 PM | Registered people only |
|
| RESTAURANT « CAFE DE LA PAIX » |
|
|
|
|
|
|
|
|
Workshop Day 1 | Wednesday June 13th |
|
|
|
|
|
|
|
|
|
| TITLES ARE TEMPORARY (except if in bold font) |
|
Registration | 08:30 |
|
|
|
|
|
Welcome and Introduction | 09:00 | Marc Snir + Franck Cappello | INRIA&UIUC | Background | Welcome, Workshop objectives and organization | |
| 09:15 | Bertrand Braunschweig | INRIA | Background | Welcome to INRIA Rennes | |
| 09:30 | Thierry Priol | INRIA | Background |
| |
Sustained Petascale | 09:45 | Bill Kramer | NCSA | Background | Blue Waters UPDATE and performance metrics | |
| 10:15 | Break |
|
|
|
|
10:45 | Romain Dolbeau | CAPS Entreprise | Background | Programming Heterogeneous Many-cores Using Directives | ||
| 11:15 | Marc Snir | ANL | Background | BlueGene Q: First impression | |
| 11:45 | Robert Ross | ANL | Background | Big Data and Scientific Computing | |
| 12:15 | Lunch |
|
|
|
|
Mini Workshop1 |
|
|
|
|
| |
Fault tolerance | 13:30 | Sanjay Kale and Marc Snir | UIUC, ANL | Background | Fault tolerance needs at NCSA and ANL | |
| 14:00 | Ana Gainaru | NCSA | Joint Result | A detailed analysis of fault prediction results and impact for HPC systems | |
| 14:30 | Amina Guermouche | INRIA | Joint Result | Unified Model for Assessing Checkpointing Protocols at Extreme-Scale | |
| 15:00 | Break |
|
|
|
|
15:30 | Mehdi Diouri | INRIA | Joint Results | Power and Energy consumption in Fault Tolerance protocols | ||
| 16:00 | Tatiana Martsinkevich | INRIA | in progress | On distributed recovery for SPMD deterministic HPC applications | |
| 16:30 | Sanjay Kale | UIUC | Result | The recovery and rise of checkpoint/restart |
|
| 17:00 | Discussions | How to address Petascale fault tolerance needs |
| ||
| 18:00 | Adjourn |
|
|
| |
Mini Workshop2 |
|
|
|
|
|
|
I/O and BigData | 13:30 | Bill Kramer and Rob Ross | UIUC, ANL | Background | I/O and BIGDATA needs at NCSA and ANL | |
| 14:00 | Matthieu Dorier | INRIA | Joint Result | In-Situ Interactive Visualization of HPC Simulations with Damaris | |
| 14:30 | Francieli Zanon Boito | UFRGS/INRIA | Joint Result | Investigating I/O approaches to improve performance and scalability of the Ocean-Land-Atmosphere Model | |
| 15:00 | Break |
|
|
|
|
15:30 | Dries Kimpe | ANL | Background | The Triton Data Model | ||
| 16:00 | --- | --- | --- |
| |
| 16:30 | --- | --- | --- |
| |
| 17:00 | Discussions | How to address Petascale I/O and Big Data needs |
| ||
| 18:00 | Adjourn |
|
|
| |
|
|
|
|
|
|
|
Workshop Day 2 | Thursday June 14th |
|
|
|
|
|
|
|
|
|
|
|
|
Math for HPC | 08:30 | Frederic Vivien | INRIA | Joint Result | Combining Process Replication and Checkpointing for Resilience on Exascale Systems | |
| 09:00 | Paul Hovland | ANL | Background | Computational Foundations of Automatic Differentiation | |
09:30 | Laurent Hascoet | INRIA | Joint Results | Gradient of MPI-parallel codes | ||
| 10:00 | Break |
|
|
|
|
Programming languages and performance modeling | 10:30 | Rajeev Thakur | ANL | Background | Recent Activities in Programming Models and Runtime Systems at ANL | |
| 11:00 | Sanjay Kale | UIUC | Background | Charj: compiler supported language with an adaptive runtime |
|
| 11:30 | Torsten Hoefler | NCSA | Background | The Sustained Petascale Performance Applications on Blue Waters Performance Considerations and Modeling | |
| 12:00 | Lunch |
|
|
| |
|
|
|
|
|
|
|
Mini Workshop3 |
|
|
|
|
|
|
Numerical libraries | 13:30 | Paul Hovland and Bill Gropp | UIUC, ANL | Background | Numerical libraries needs at NCSA and ANL | |
| 14:00 | Laura Grigori | INRIA | Joint Result | Hybrid static/dynamic scheduling for already optimized dense matrix factorization | |
| 14:30 | François Pelegrini | INRIA | Joint Result | Introducing PaMPA | |
| 15:00 | Break |
|
|
|
|
| 15:30 | Jocelyne Erhel | INRIA | Background | Solving linear systems arising from flow simulations in 3D Discrete Fracture Networks |
|
16:00 | Daisuke Takahashi | U. Tsukuba | Joint Result | An Implementation of Parallel 3-D FFT with 1.5-D Decomposition | ||
| 16h30 | Adrien Remy | INRIA | Joint Result | Solving general dense linear systems on hybrid multicore-GPU systems. | |
| 17:00 | Discussions | How to address Petascale Numerical Libraries needs |
| ||
| 17:55 | Adjourn |
|
|
| |
|
|
|
|
|
|
|
Mini Workshop4 |
|
|
|
|
|
|
Programing Models | 13:30 | Rajeev Thakur and Sanjay Kale | UIUC, ANL | Background | Programming model needs at NCSA and ANL | |
| 14:00 | Jean-François Mehaut | INRIA | Joint Result | Load Balancing for Parallel Multi-core Machines with Non-Uniform Communication Costs | |
| 14:30 | Brice Goglin | INRIA | Background | Bringing hardware affinity information into MPI communication strategies | |
| 15:00 | Break |
|
|
|
|
| 15:30 | Thomas Ropars | EPFL | Background | Towards efficient collective operations on the Intel SCC | |
16:00 | Alexandre Duchateau | INRIA | Joint Result | Hydra : Generation and Tuning of Parallel Solutions for Linear Algebra | ||
| 16:30 | Discussions | How to address Petascale programing model needs |
| ||
| 17:30 | Adjourn |
|
|
| |
|
|
|
|
|
|
|
| 18:00 | Banquet |
|
| @ Saint Malo |
|
|
|
|
|
|
|
|
Workshop Day 3 | Friday June 15th |
|
|
|
|
|
|
|
|
|
|
|
|
Mini Workshop5 |
|
|
|
|
|
|
Mapping and Scheduling | 08:30 | BIll Kramer and Marc Snir | UIUC, ANL | Background | Mapping and Scheduling needs at NCSA and ANL | |
| 09:00 | François Teyssier | INRIA | Joint Result | Load balacing and affinities between processes with TreeMatch in Charm++ : preliminary results and prospects |
|
| 09:30 | Sébastien Fourestier | INRIA | Background | Latest improvements to Scotch and ongoing collaborations |
|
| 10:00 | Torsten Hoefler | NCSA --> ETH | Background | On-node and off-node Topology Mapping for Petascale Computers |
|
| 10:30 | Joseph Emeras, Olivier Richard, Cristian Ruiz | INRIA | Background | Jobs Resource Utilization as a Metric for Clusters Comparison and Optimization |
|
| 11:00 | Discussions | How to address Petascale Mapping and Scheduling needs |
| ||
Mini Workshop6 |
|
|
|
|
|
|
HPC/Cloud | 08:30 | Kate Keahey (main speaker) | ANL, INRIA | Background | HPC Cloud | |
| 09:00 | Gabriel Antoniu | INRIA | Joint Result | A Performance Evaluation of Azure and Nimbus Clouds for Scientific Applications |
|
| 09:30 | Frederic Desprez | INRIA | Background | Budget Constrained Resource Allocation for Non-Deterministic Workflows on a IaaS Cloud |
|
| 10:00 | Bogdan Nicolae | INRIA | Joint Results | A Hybrid Local Storage Transfer Scheme for Live Migration of I/O Intensive Workloads |
|
| 10:30 | Derrick Kondo | INRIA | Result | Characterization and Prediction of Host Load in a Google Data Center |
|
| 11:00 | Discussions | How to address HPC Cloud needs |
| ||
|
|
|
|
|
|
|
| 12:00 | Franck Cappello and Marc Snir |
|
| Discussion and Closing |
|
|
|
|
|
|
|
|
| 12:30 | Lunch |
|
|
|
|
Abstracts
Romain Dolbeau: Programming Heterogeneous Many-cores Using Directives
...
We highlight different algorithms for solving general dense linear systems using LU decomposition on multicore systems accelerated with multiple GPUs. We present a set of techniques to achieve high efficiency on these architectures and we compare the resulting algorithms in terms of performance, speed-up and accuracy. Finally we present some ongoing work with the goal to move to a larger scale using clusters of GPUs.
*Derrick Kondo: Characterization and Prediction of Host Load in a Google Data Center*
\ Wiki Markup
[Joint work with Sheng Di, Walfredo Cirne\]
Characterization and prediction of system load is essential for optimizing its performance. We characterize and predict host load in a real-world production Google data center, using a detailed trace of over 25 million tasks across over 12,500 hosts. In our characterization, we present valuable statistics and distributions of machine’s maximum load, queue state and relative usage levels. Compared to traditional load from Grids and other HPC systems, we ?nd that Google host load exhibits higher variance due to the high frequency of small tasks. Based on this characterization, we develop and evaluate different methods of machine load prediction using techniques such as autoregression, moving averages, probabilistic methods, and noise ?lters. We ?nd that a linear weighted moving average produces accurate predictions with a 80%-95% success rate, outperforming other methods by 5%-20%. Surprisingly, this method outperforms more sophisticated hybrid prediction methods, which are effective for traditional Grid loads but not data center loads due to its more frequent and severe ?uctuations.
...