Programme

Full Proceedings now available for download here

Day 1, 3 October 2013

 

09:00 - 09:10 Welcome address - Prof Mark Parsons, EPCC Executive Director
09:10 - 10:10

Keynote: Efficient hardware support for PGAS languages

Duncan Roweth (Cray)

10:10 - 10:30 COFFEE BREAK
  Research Session 1
10:30 - 11:00

 

Optimizing Collective Communication in OpenSHMEM

Jithin Jose, Krishna Chaitanya Kandalla, Sreeram Potluri,
Jie Zhang and Dhabaleswar Panda (Ohio State University)

11:00 - 11:30

Productivity and Performance of the HPC Challenge Benchmarks
with the XcalableMP PGAS language

Masahiro Nakao (University of Tsukuba), Hiroshi Murai,
Takenori Shimosaka (RIKEN Advanced Institute for Computational Science) and
Mitsuhisa Sato (University of Tsukuba)

11:30 - 12:00

Additional Coarray Features in Fortran

John Reid (JKR Associates)

12:00 - 12:30

Introducing SHMEM into the GROMACS molecular dynamics application:
experience and results

Ruyman Reyes-Castro, Andrew Turner (EPCC) and
Berk Hess (KTH Royal Institute of Technology)

12:30 - 13:30 LUNCH
  Hot Session 1
13:30 - 13:50

UPCGAP: A UPC package for the GAP system

Nick Johnson (EPCC), Alexander Konovalov, Vladimir Janjic and
Steve Linton (University of St Andrews)

13:50 - 14:10

Evaluation of Unified Parallel C for Molecular Dynamics

Kamran Idrees, Christoph Niethammer (HLRS), Aniello Esposito (Cray) and
Colin W. Glass (HLRS)

14:10 - 14:30

First Experiences on Collective Operations with Fortran Coarrays on the Cray XC30

Pekka Manninen and Harvey Richardson (Cray)

  Research Session 2
14:30 - 15:00

A scalable deadlock detection algorithm for UPC collective operations

Indranil Roy, Glenn Luecke, James Coyle and Marina Kraeva (Iowa State University)

15:00 - 15:30

An Efficient Implementation of Stencil Communication for the
XcalableMP PGAS Parallel Programming Language

Hitoshi Murai (RIKEN Advanced Institute for Computational Science) and
Mitsuhisa Sato (University of Tsukuba)

15:30 - 16:00 COFFEE BREAK
  Hot Session 2
16:00 - 16:20

Asynchronous Global Heap: Stepping Stone to Global Memory Management

Yuichiro Ajima, Hideyuki Akimoto, Tomoya Adachi, Takayuki Okamoto, Kazushige Saga, Kenichi Miura and Shinji Sumimoto (Fujitsu Limited)

16:20 - 16:40

Static Analyses for Unaligned Collective Synchronization Matching for OpenSHMEM

Swaroop Pophale (University of Houston), Oscar Hernandez, Stephen Poole (Oak Ridge National Laboratory) and Barbara Chapman (University of Houston)

16:40 - 17:00

Improving Performance of All-to-All Communication Through Loop Scheduling in UPC

Michail Alvanos (Barcelona Supercomputing Center), Ilie Tanase (IBM), Montse Farreras (Universitat Politècnica de Catalunya), Ettore Tiotto (IBM), Jose Nelson (University of Alberta) and Xavier Martorell (Universitat Politècnica de Catalunya)

  Research Session 3
17:00 - 17:30

Performance Analysis of a Large Memory Application on Multiple Architectures

Peter Kogge (University of Notre Dame)

 
17:30 - 19:30

Poster & Drinks Reception

A chance to look at the Posters, displayed in the Duddingston Room
and to chat to fellow conference attendees over a few drinks and nibbles

Posters

The performance of the PCJ library for massively parallel computing 
Marek Nowicki (Nicolaus Copernicus University Torun) and
Piotr Bala (ICM, University of Warsaw)

Parallel Algebraic Multigrid Coarsening Using PGAS Languages
Niclas Jansson (KTH Royal Institute of Technology)

A Comparison of PGAS Languages on Multi-core Clusters
Konstantina Panagiotopoulou and Hans-Wolfgang Loidl (Heriot-Watt University)

Communication and Computation Overlapping Parallel Fast Fourier Transform in Cray UPC
Erik Jarleberg, Johannes Bulin (KTH Royal Institute of Technology), Jens Doleschal (TU Dresden), Stefano Markidis and Erwin Laure (KTH Royal Institute of Technology)

Validation and Verification Suite for OpenSHMEM
Swaroop Pophale, Tony Curtis, Barbara Chapman (University of Houston)
and Stephen Poole (Oak Ridge National Laboratory)

 

 

Day 2, 4 October 2013

09:00 - 10:00

Keynote: Issues for Exascale Accelerated Computing - system
architecture and programming

Mitsuhisa Sato (University of Tsukuba)

10:00 - 10:20 COFFEE BREAK
  Research Session 4
10:20 - 10:50

Fortran coarray library for 3D cellular automata microstructure simulation

Anton Shterenlikht (University of Bristol)

10:50 - 11:20

Coarray C++

Troy Johnson (Cray) and David Henty (EPCC)

11:20 - 11:50

On the Modelling of One-Sided Communication Systems

Olaf Krzikalla, Andreas Knuepfer, Ralph Mueller-Pfefferkorn and
Wolfgang E. Nagel (ZIH, TU Dresden)

11:50 - 12:20

Flat Combining Synchronized Data Structures

Brandon Holt, Jacob Nelson, Brandon Myers, Preston Briggs, Luis Ceze,
Simon Kahan and Mark Oskin (University of Washington)

12:20 - 13:10 LUNCH
  Hot Session 3
13:10 - 13:30

A PGAS infrastructure for Heterogeneous Reconfigurable Computing

Ruediger Willenberg and Paul Chow (University of Toronto)

13:30 - 13:50

Model Checking with User-Definable Memory Consistency Models

Tatsuya Abe and Toshiyuki Maeda (RIKEN Advanced Institute for
Computational Science)

13:50 - 14:10

The GASPI API specification and its implementation GPI 2.0

Daniel Grünewald (Fraunhofer ITWM) and Christian Simmendinger (T-Systems)

  Research Session 5
14:10 - 14:40

Applying Type Oriented Programming to the PGAS memory model

Nick Brown (EPCC)

14:40 - 15:10

UPC on MIC: Early Experiences with Native and Symmetric Modes

Miao Luo, Mingzhe Li, Akshay Venkatesh, Xiaoyi Lu, Sreeram Potluri and
Dhabaleswar Panda (Ohio State University)

15:10 - 15:30 COFFEE BREAK
  Research Session 6
15:30 - 16:00

PGAS implementation of SpMVM and LBM using GPI

Faisal Shahzad, Markus Wittmann, Moritz Kreutzer, Thomas Zeiser, Georg Hager (Friedrich-Alexander-Universität Erlangen-Nürnberg) and
Gerhard Wellein (Erlange Regional Computing Centre)

16:00 - 16:30

Guiding X10 Programmers to Improve Runtime Performance

Jeeva Paudel (University of Alberta), Olivier Tardieu (IBM) and
Jose Nelson Amaral (University of Alberta)

16:30 - 17:00

Programming Large Dynamic Data Structures on a DSM Cluster of Multicores

Sai Charan Koduru, Min Feng and Rajiv Gupta (University of California Riverside)

 
17:00 - 17:15 Conference wrap-up & PGAS2014