SBAC-PAD 2018 Program (pdf)

 

Monday, September 24, 2018

08:00-17:00
Ecole Normale Superieure of Lyon - Site Buisson - Main Entrance: Registration
 
Workshops & Tutorials Day
Location
D8 001 Room
09:00-10:00

Joint Keynote 1: "Serendipity: How Supercomputing Technology is Enabling
a Revolution in Artificial Intelligence ",
José E. Moreira, IBM Research

Keynote chair: Laurent Lefevre

Abstract

With the availability of both large compute power and large data sets, we have witnessed a revolution in machine learning technology, which has become a mainstream tool for both business and scientific applications. This revolution is likely to accelerate, as even more compute power is brought to bear, and deliver many of the promises of artificial intelligence. In this talk we will investigate how far the impacts of machine learning can go. We will cover the new Summit supercomputer, which brings unprecedented compute capabilities to both traditional high performance computing and artificial intelligence problems, analyzing the similarities as well as the differences in those two fields. We will also speculate about the future of machine learning and, in particular, its possible limitations. We will conclude with a discussion of one of the most important scientific questions of our time: Is consciousness computable?


Jose MoreiraJosé E. Moreira is a Distinguished Research Staff Member in the Scalable Systems Department at the Thomas J. Watson Research Center. He received a B.S. degree in physics and B.S. and M.S. degrees in electrical engineering from the University of Sao Paulo, Brazil, in 1987, 1988 and 1990, respectively. He also received a Ph.D. degree in electrical engineering from the University of Illinois at Urbana-Champaign in 1995. Since joining IBM at the Thomas J. Watson Research Center, he has worked on a variety of high-performance computing projects. He was system software architect for the Blue Gene/L supercomputer and chief architect of the Commercial Scale Out project. He currently leads the IBM Research work on the architecture of Power processor. He is an author or coauthor of over 100 technical papers and 10 patents. Dr. Moreira is a member of the IEEE (Institute of Electrical and Electronics Engineers) and a Distinguished Scientist of the ACM (Association for Computing Machinery).

10:00-10:30
Coffee & Tea & Juice Break
Location
Room Buisson 1
D8 001
Room Buisson 2
D8 003
Room Buisson 3
D8 006
Room Buisson 4
D8 007
10:30-13:00 HPLM Workshop WAMCA Workshop Tutorial 1: Secure execution in the cloud using Intel SGX Tutorial 2: Singularity: an HPC application container
13:00-14:00 Lunch at the Restaurant ENS Descartes
14:00-16:00 HPLM Workshop WAMCA Workshop Tutorial 3: Distributed computing experiments using Grid’5000 / SILECS testbed 
16:00-16:30 Coffee & Tea & Juice Break
16:30-18:00 HPLM Workshop WAMCA Workshop Tutorial 3: Distributed computing experiments using Grid’5000 / SILECS testbed 

 

 

Tuesday, September 25, 2018

08:00-17:00
Ecole Normale Superieure of Lyon - Site Buisson - Main Entrance: Registration
09:00-09:15
Room Buisson 1 - D8 001 : SBAC-PAD 2018 - Day 1 - Opening Session
Laurent Lefèvre, Marcos Dias de Assuncao, Alfredo Goldman, Lucas Schnorr
09:15-10:15
Keynote 2: "Scheduling Matters", Yves Robert, ENS Lyon, France and Univ. Tenn. Knoxville, USA
Keynote chair: Alfredo Goldman

Abstract

This talk will review a few scheduling algorithms to solve simple computational problems on large-scale platforms. Faults, energy/power shortage, I/O contention, the constraints are numerous and challenging. The talk will provide a few answers and discuss open research directions.


Yves Robert Yves Robert received the PhD degree from Institut National Polytechnique de Grenoble. He is currently a full professor in the Computer Science Laboratory LIP at ENS Lyon. He is the author of 7 books, 150 papers published in international journals, and 240 papers published in international conferences. He is the editor of 11 book proceedings and 13 journal special issues. He has advised 30 PhD students. His main research interests are scheduling techniques and resilient algorithms for large-scale platforms. Yves Robert served on many editorial boards, including IEEE TPDS, JPDC and ACM TOPC. He is a Fellow of the IEEE. He was elected a Senior Member of Institut Universitaire de France in 2007 and renewed in 2012. He was awarded the 2014 IEEE TCSC Award for Excellence in Scalable Computing, and the 2016 IEEE TCPP Outstanding Service Award. He holds a Visiting Scientist position at the University of Tennessee Knoxville since 2011.
10:15-10:45
Coffee & Tea & Juice Break
 

Conference

Location
D8 001
10:45-12:30
Session 1: Computer Architecture and Compilers
Session chair: Philippe Olivier Alexandre Navaux (UFRGS)
  • MLNoC: A Machine Learning Based Approach to Network-on-Chip Design:  Nishant Rao, Akshay Ramachandran and Amish Shah
  • ADeLe: Rapid Architectural Simulation for Approximate Hardware: Isaias Bittencourt Felzmann, Matheus M. Susin, Liana Duenha, Rodolfo Azevedo and Lucas Wanner
  • From Java to FPGA: an Experience with the Intel HARP System: Pedro Caldeira, Jerônimo Penha, Lucas Bragança, José Nacif, Ricardo Ferreira, Renato Ferreira and Fernando Quintao
  • Online Detection of Spectre Attacks Using Microarchitectural Traces from Performance Counters: Congmiao Li and Jean-Luc Gaudiot (short paper)
  • DOACROSS Parallelization based on Component Annotation and Loop-carried Probability: Luis Felipe Mattos, César Divino, Juan Salamanca, Joao Paulo Carvalho, Marcio Machado Pereira and Guido Araujo (short paper)

12:30-14:00
Lunch at the Restaurant ENS Descartes
14:00-16:00
Session 2: Scheduling
Session chair: Arnaud Legrand (CNRS)
  • Scheduling independent stochastic tasks under deadline and budget constraints: Louis-Claude Canon, Aurélie Kong Win Chang, Yves Robert and Frédéric Vivien
  • Adaptive scheduling of collocated applications using a task-based runtime system: Jiri Dokulil and Siegfried Benkner
  • A Batch Task Migration Approach for Decentralized Global Rescheduling: Vinicius Freitas, Alexandre Santana, Márcio Castro and Laércio L. Pilla
  • Exploring Power Budget Scheduling Opportunities and Trade-offs for AMR-based Applications: Yubo Qin, Ivan Rodero, Pradeep Subedi, Manish Parashar and Sandro Rigo
  • EASE: Energy Efficiency and ProportionalityAware Virtual Machine Scheduling: Congfeng Jiang, Yumei Wang, Dongyang Ou, Yeliang Qiu, Youhuizi Li, Jian Wan, Bing Luo, Weisong Shi and Christophe Cerin (short paper)
16:00-16:30
Coffee & Tea & Juice Break
16:30-18:00
 Session 3: Energy in the Cloud, Network
Session chair: Ivan Rodero (Rutgers University)
  • Energy-Efficient IaaS-PaaS Co-design for Flexible Cloud Deployment of Scientific Applications: David Guyon, Anne-Cécile Orgerie and Christine Morin
  • Frequency Selection Approach for Energy Aware Cloud Database: Chaopeng Guo and Jean-Marc Pierson
  • Network-aware energy-efficient virtual machine management in distributed Cloud infrastructures with on-site photovoltaic production: Benjamin Camus, Fanny Dufossé, Anne Blavette, Martin Quinson and Anne-Cécile Orgerie
  • A Novel Broker-Based Hierarchical Authentication Scheme in Proxy Mobile IPv6 Networks: Jang Su Hwan and Jeong Jongpil (short paper)
18:30-20:30 Welcome Reception and Cocktail at the Restaurant ENS Descartes and Buisson Garden

 

 

 

Wednesday, September 26, 2018

08:00-17:00
Ecole Normale Superieure of Lyon - Site Buisson - Main Entrance: Registration
09:00-09:15
Room Buisson 1 - D8 001 : SBAC-PAD 2018  - Day 2 -Opening Session
Laurent Lefèvre, Marcos Dias de Assuncao, Alfredo Goldman, Lucas Schnorr
09:15-10:15
Keynote 3: "Extreme-Scale Earthquake Simulation on Sunway TaihuLight", Haohuan Fu, Tsinghua University, China
Keynote chair: Lucas Schnorr

Abstract

This talk would first introduce and discuss the design philosophy about the Sunway TaihuLight system, and then describe our recent efforts on performing earthquake simulations on such a large-scale system. Our work in 2017 accomplished a complete redesign of AWP-ODC for Sunway architectures, achieves over 15% of the system's peak, better than the 11.8% achieved by a similar software running on Titan, whose byte to flop ratio is 5 times better than TaihuLight. The extreme cases demonstrate a sustained performance of over 18.9 Pflops, enabling the simulation of Tangshan earthquake as an 18-Hz scenario with an 8-meter resolution. Our recent work further improves the simulation framework with capabilities to describe complex surface topography, and to drive building damage prediction and landslide simulation, which are demonstrated with a case study of the Wenchuan earthquake with accurate surface topography and improved coda wave effects.


Haohuan FuHaohuan Fu is the deputy director of the National Supercomputing Center in Wuxi, leading the research and development division. He is also an associate professor in the Ministry of Education Key Laboratory for Earth System Modeling, and Department of Earth System Science in Tsinghua University, where he leads the research group of High Performance Geo-Computing (HPGC). Fu has a PhD in computing from Imperial College London. Since joining Tsinghua in 2011, Dr. Fu has been working towards the goal of providing both the most efficient simulation platforms and the most intelligent data management and analysis platforms for geoscience applications. His research has, for example, led to efficient designs of atmospheric dynamic solvers for both Tianhe-1A, Tianhe-2, Sunway TaihuLight supercomputers, and the reconfigurable computing platforms. The work based on the Sunway TaihuLight supercomputer manages to scale a fully-implicit solver to over 10 million cores, which won the Gordon Bell Prize of SC16.

10:15-10:45
Coffee & Tea & Juice Break
 

Conference

Location
D8 001
10:45-12:30
Session 4: Applications
Session chair: Guido Araujo (UNICAMP)
  • Designing a Parallel Memory-Aware Lattice Boltzmann Algorithm on Manycore Systems: Yuankun Fu, Feng Li, Fengguang Song and Luoding Zhu
  • A new efficient parallel algorithm for minimum spanning tree: Jucele Vasconcellos, Edson Caceres, Henrique Mongelli and Siang Song
  • Exploring Self-Adaptivity towards Performance and Energy for Time-stepping Methods: Natalia Kalinnik, Robert Kiesel, Thomas Rauber, Marcel Richter and Gudula Rünger
  • Predicting the Reliability Behavior of HPC Applications: Daniel Oliveira, Francis Birck Moreira, Paolo Rech and Philippe Navaux
12:30-13:30
Lunch at the Restaurant ENS Descartes
13:30-15:30
Session 5: GPU based computing
Session chair: Eduardo Rodrigues (IBM Research)
  • Variable-size batched condition number calculation on GPUs: Hartwig Anzt, Jack Dongarra, Goran Flegar and Thomas Grutzmacher
  • Towards a Single-host Many-GPU System: Ming-Hung Chen, Ihsin Chung, Bulent Abali and Paul Crumley
  • Exploiting Limited Access Distance for Kernel Fusion Across the Stages of Explicit One-Step Methods on GPUs: Tim Werner and Matthias Korch
  • Balancing load of GPU subsystems to accelerate image reconstruction in parallel beam tomography: Suren Chilingaryan, Evelina Ametova, Andreas Kopmann and Alessandro Mirone
  • Performance Prediction of GPU-based Deep Learning Applications: Eugenio Gianniti, Li Zhang and Danilo Ardagna (short paper)
15:30-16:30
Poster Session during the Coffee & Tea & Juice Break
Posters

 

  • Verlet buffer for broad phase interaction detection in Discrete Element Method: Abdoul Wahid Mainassara Chekaraou, Alban Rousset, Xavier Besseron and Bernhard Peters
  • Deep Reinforcement Learning Based Approach for Energy-Aware Job Scheduling in Grid Computing: Lucas Casagrande, Kleiton Pereira, Renato Tanaka, Charles Miers, Guilherme Koslovski and Mauricio Pillon
  • Analysis and Characterization of Control Network Traffic in OpenStack Based IaaS Clouds: Tiago Reinert, Adnei Donatti, Guilherme Koslovski, Maurício Pillon and Charles Miers
  • Performance and Energy Consumption of Parallel Programming Interfaces in Multicore Architectures: A Case Study: Adriano Garcia, Claudio Schepke, Alessandro Girardi, Sherlon Almeida Da Silva
  • Automatic Parallelization for Shared Memory Scientific Multiprocessing: A State-of-the-art: Re'Em Harel, Idan Mosseri, Harel Levin, Matan Rusanovsky, Gal Oren
16:30-18:00
 Session 6: Programming Paradigms and Memory
Session chair: Nelson J. Amaral (UAlberta)
  • Polyhedral Dataflow Programming: a Case Study: Romain Fontaine, Laure Gonnord and Lionel Morel
  • Enabling Efficient Job Dispatching in Accelerator-extended Heterogeneous Systems with Unified Address Space: Georgios Kornaros
  • Phase-Based Data Placement Scheme for Heterogeneous Memory Systems: Mohammad Laghari, Najeeb Ahmad and Didem Unat
  • Exploiting Compute Caches for Memory Bound Vector Operations: Joao Vieira, Paolo Ienne, Nuno Roma, Gabriel Falcao and Pedro Tomas (short paper)
 

Free evening

 

 

 

Thursday, September 27, 2018

08:00-17:00
Ecole Normale Superieure of Lyon - Site Buisson - Main Entrance: Registration
09:00-09:15
Room Buisson 1 - D8 001: SBAC-PAD 2018 - Day 3 - Opening Session
Laurent Lefèvre, Marcos Dias de Assuncao, Alfredo Goldman, Lucas Schnorr
09:15-10:15
Keynote 4: "Big Data at Extreme-Scales: Addressing Computational
Challenges in the 21st Century", Manish Parashar, Rutgers University, USA 

Keynote chair: Laurent Lefevre

Abstract

Data-related challenges are quickly dominating computational and data-enabled sciences and are limiting the potential impact of scientific application workflows enabled by current and emerging extreme scale, high-performance, distributed computing environments. These data-intensive application workflows involve dynamic coordination, interactions and data coupling between multiple application processes that run at scale on different resources, and with services for monitoring, analysis and visualization and archiving, and present challenges due to increasing data volumes and complex data-coupling patterns, system energy constraints, increasing failure rates, etc. In this talk I will explore some of these challenges and investigate how solutions based on data sharing abstractions, managed data pipelines, data-staging service, and in-situ / in-transit data placement and processing can be used to help address them. This research is part of the DataSpaces project at the Rutgers Discovery Informatics Institute.


Manish ParasharManish Parashar is Distinguished Professor of Computer Science at Rutgers University. He is also the founding Director of the Rutgers Discovery Informatics Institute (RDI2). His research interests are in the broad areas of Parallel and Distributed Computing and Computational and Data-Enabled Science and Engineering. Manish is the founding chair of the IEEE Technical Consortium on High Performance Computing (TCHPC), Editor-in-Chief of the IEEE Transactions on Parallel and Distributed Systems, and serves on the editorial boards and organizing committees of a large number of journals and international conferences and workshops. He has over 350 publications, has deployed several software systems that are widely used, and has received a number of awards for his research and leadership. Manish is Fellow of AAAS, Fellow of IEEE/IEEE Computer Society and ACM Distinguished Scientist.

10:15-10:45
Coffee & Tea & Juice Break
 

Conference

Location
D8 001
10:45-12:30
Session 7: Data Analytics, Locality and I/O
Session chair:Cristiana Bentes (UERJ)
 
  • Exploring the Potential of Next Generation Software-Defined In-Memory Frameworks: Shouwei Chen and Ivan Rodero
  • Towards Green Scientific Data Compression Through High-Level I/O Interfaces: Yevhen Alforov, Anastasiia Novikova, Michael Kuhn, Julian Kunkel and Thomas Ludwig
  • Improving Data Locality in P2P-based Fog Computing Platforms: Luiz Angelo Steffenel
  • echofs: A Scheduler-guided Temporary Filesystem to leverage Node-local NVMs: Alberto Miranda, Ramon Nou and Toni Cortes (short paper)
  • A Jaccard Weights Kernel Leveraging Independent Thread Scheduling on GPUs: Hartwig Anzt and Jack Dongarra (short paper)
12:30-14:00
Lunch at the Restaurant ENS Descartes
14:00-16:00
Session 8:  Performance Prediction and Evaluation
Session chair: Hartwig Anzt (University of Tennessee)
  • Multicore Performance Engineering of Sparse Triangular Solves Using a Modified Roofline Model: Markus Wittmann, Georg Hager, Radim Janal ́ık, Martin Lanser, Axel Klawonn, Oliver Rheinbach, Olaf Schenk and Gerhard Wellein
  • Predicting the Performance Impact of Increasing Memory Bandwidth for Scientific Workflows: Nelson Mimura Gonzalez, Jose Brunheroto, Fausto Artico, Yoonho Park, Tereza Carvalho, Charles Christian Miers, Mauricio Aronne Pillon and Guilherme Piegas Koslovski
  • Mainstream vs. Emerging HPC: Metrics, Trade-offs and Lessons Learned: Milan Radulovic, Kazi Asifuzzaman, Darko Zivanovic, Nikola Rajovic, Guillaume Colin de Verdi`ere, Dirk Pleiter, Manolis Marazakis, Nikolaos Kallimanis, Paul Carpenter, Petar Radojkovic and Eduard Ayguadé
  • Assessing Time Predictability Features of ARM big.LITTLE Multicores: Gabriel Fernandez, Francisco J Cazorla, Jaume Abella and Sylvain Girbal (short paper)
  • Adaptive Partitioning for Iterated Sequences of Irregular OpenCL Kernels: Pierre Huchant, Marie-Christine Counilh and Denis Barthou (short paper)
16:00-16:30
Coffee & Tea & Juice Break
16:30-18:00
Session 9: IoT, Fog, Edge, and Cloud Computing
Session chair: Anne-Cécile Orgerie (CNRS)
  • Partitioning convolutional neural networks for inference on constrained Internet-of-Things devices: Fabiola Martins Campos de Oliveira and Edson Borin
  • Runtime Management of Data Quality for Scientific Observatories Using Edge and In-Transit Resources: Ali Reza Zamani, Daniel Balouek-Thomert, J. J. Villalobos, Ivan Rodero and Manish Parashar
  • A Fault-Tolerant Agent-based Architecture for Transient Servers in Fog Computing: Jose Pergentino de Araujo Neto, Célia G. Ralha and Donald M. Pianto
19:00-22:30

 

Banquet Reception on the Hermes boat / 3 Hours Cruise along the Rhone and Saone Rivers

 

Hermes Buffet  Hermes Buffet

 

Best Paper Award Ceremony

Hermes

 

 

Annoucement of SBAC-PAD 2019

 

Hermes Buffet Hermes Buffet