Workshop on Latest Advances in Scalable Algorithms for Large-Scale Systems (ScalA)
Novel scalable scientific algorithms are needed in order to enable key science
applications to exploit the computational power of large-scale systems. This is
especially true for the current tier of leading petascale machines and the road
to exascale computing as HPC systems continue to scale up in compute node and
processor core count. These extreme-scale systems require novel scientific
algorithms to hide network and memory latency, have very high
computation/communication overlap, have minimal communication, and have no
synchronization points.
Scientific algorithms for multi-petaflop and exa-flop systems also need to be
fault tolerant and fault resilient, since the probability of faults increases
with scale. Resilience at the system software and at the algorithmic level is
needed as a crosscutting effort. Finally, with the advent of heterogeneous
compute nodes that employ standard processors as well as GPGPUs, scientific
algorithms need to match these architectures to extract the most performance.
This includes different system-specific levels of parallelism as well as
co-scheduling of computation. Key science applications require novel mathematical
models and system software that address the scalability and resilience challenges
of current- and future-generation extreme-scale HPC systems.
Submission Guidelines
Authors are invited to submit manuscripts in English structured as
technical papers not exceeding 8 letter size (8.5x11) pages including
figures, tables, and references using the IEEE format for conference
proceedings. Submissions not conforming to these guidelines may be
returned without review. Reference style files are available at
http://www.ieee.org/conferences_events/conferences/publishing/templates.html.
All manuscripts will be reviewed and judged
on correctness, originality, technical strength, and significance,
quality of presentation, and interest and relevance to the workshop
attendees. Submitted papers must represent original unpublished research
that is not currently under review for any other conference or journal.
Papers not following these guidelines will be rejected without review
and further action may be taken, including (but not limited to)
notifications sent to the heads of the institutions of the authors and
sponsors of the conference. Submissions received after the due date,
exceeding length limit, or not appropriately structured may also not be
considered. At least one author of an accepted paper must register for
and attend the workshop. Authors may contact the workshop program chair
for more information.
Papers should be submitted electronically at: https://www.easychair.org/conferences/?conf=scala20140.
Full papers will be published with the SC'14 workshop proceedings in
the IEEE and ACM digital libraries. Selected papers will be invited for an extended
version in a special issue of the Journal of Computational Science (JoCS).
Important Dates
- Full paper submission:
7 September, 2014
- Notification of acceptance:
23 September, 2014
- Final paper submission (firm):
6 October, 2014
Topics
Topics of interest include, but are not limited to:
- Novel scientific algorithms that improve performance, scalability,
resilience, and power efficiency
- Porting scientific algorithms and applications to many-core and
heterogeneous architectures
- Performance and resilience limitations of scientific algorithms and
applications at scale
- Crosscutting approaches (system software and applications) in
addressing scalability challenges
- Scientific algorithms that can exploit extreme concurrency
(e.g. 1 billion for exascale by 2020)
- Naturally fault tolerant, self-healing, or fault oblivious scientific
algorithms
- Programming model and system software support for algorithm
scalability and resilience
Workshop Chairs
- Prof. Vassil Alexandrov, Barcelona Supercomputing Center, Spain
- Al Geist, Oak Ridge National Laboratory, USA
Workshop Program Chair
- Dr. Christian Engelmann, Oak Ridge National Laboratory, USA
Program Committee
- Prof. Vassil Alexandrov, Barcelona Supercomputing Center, Spain
- Dr. Rick Archibald, Oak Ridge National Laboratory, USA
- Dr. David E. Bernholdt, Oak Ridge National Laboratory, USA
- Dr. Greg Bronevetsky, Lawrence Livermore National Laboratory, USA
- Dr. Michael Heroux, Sandia National Laboratories
- Dr. Mark Hoemmen, Sandia National Laboratories
- Prof. Marian Bubak, AGH University of Science and Technology,
Krakow, Poland and University of Amsterdam, The Netherlands
- Prof. Zizhong Chen, University of California, Riverside, USA
- Dr. Christian Engelmann, Oak Ridge National Laboratory, USA
- Dr. Kirk E. Jordan, IBM T.J. Watson Research, USA
- Prof. Dieter Kranzlmueller, Ludwig-Maximilians-University Munich, Germany
- Prof. Ron Perrot, University of Oxford, UK
- Dr. Nageswara Rao, Oak Ridge National Laboratory, USA
Venue
- Room 283-84-85, New Orleans Ernest N. Morial Convention Center, 900 Convention Center Blvd, New Orleans, LA 70130, USA
Program
- 09:00-10:05 Session 1
- 09:00-09:05 Opening
- 09:05-09:45 Keynote 1: "Performance Analytics for Large Scale Computations,"
Jesus Labarta (Barcelona Supercomputing Center, Spain)
(Abstract)
- 09:45-10:05 Paper 1: "Scaling Parallel 3-D FFT with Non-Blocking MPI Collectives,"
Sukhyun Song and Jeffrey K. Hollingsworth
- 10:05-10:30 Coffee break (coffee provided)
- 10:30-12:10 Session 2
- 10:30-11:10 Keynote 2: "Fault Tolerance in Numerical Library Routines,"
Jack Dongarra (University of Tennessee, Knoxville, USA)
(Abstract)
- 11:10-11:30 Paper 2: "Exploiting Data Representation for Fault Tolerance,"
James Elliott, Mark Hoemmen and Frank Mueller
- 11:30-11:50 Paper 3: "VCube: A Provably Scalable Distributed Diagnosis Algorithm,"
Elias P. Duarte Jr., Luis C.E. Bona and Vinicius K. Ruoso
- 11:50-12:10 Paper 4: "TX: Algorithmic Energy Saving for Distributed Dense Matrix Factorizations,"
Li Tan and Zizhong Chen (presented by Panruo Wu)
- 12:10-13:30 Lunch break (lunch on your own)
- 13:30-15:10 Session 3
- 13:30-14:10 Keynote 3: "Unite and Conquer Approach for Large Scale Numerical Computing,"
Nahid Emad (University of Versailles, France)
(Abstract)
- 14:10-14:30 Paper 5: "CholeskyQR2: A Simple and Communication-Avoiding Algorithm for Computing a Tall-Skinny QR Factorization on a Large-Scale Parallel System,"
Takeshi Fukaya, Yuji Nakatsukasa, Yuka Yanagisawa and Yusaku Yamamoto
- 14:30-14:50 Paper 6: "Deflation Strategies to Improve the Convergence of Communication-Avoiding GMRES,"
Ichitaro Yamazaki, Stanimire Tomov, and Jack Dongarra
- 14:50-15:10 Paper 7: "A Framework for Parallel Genetic Algorithms for Distributed Memory Architectures,"
Dobromir Georgiev, Emanouil Atanassov and Vassil Alexandrov
- 15:10-15:30 Coffee break (coffee provided)
- 15:30-17:30 Session 4
- 15:30-16:10 Keynote 4: "Dynamic Big Data Applications,"
Craig C. Douglas (University of Wyoming, USA)
(Abstract)
- 16:10-16:30 Paper 8: "The Anatomy of Mr. Scan: A Dissection of Performance of an Extreme Scale GPU-Based Clustering Algorithm,"
Benjamin Welton and Barton Miller
- 16:30-16:50 Paper 9: "Performance and Portability with OpenCL for Throughput-Oriented HPC Workloads Across Accelerators, Coprocessors, and Multicore Processors,"
Chongxiao Cao, Mark Gates, Azzam Haidar, Piotr Luszczek, Stanimire Tomov, Ichitaro Yamazaki and Jack Dongarra
- 16:50-17:10 Paper 10: "A Hierarchical Tridiagonal System Solver for Heterogenous Supercomputers,"
Xinliang Wang, Yangtong Xu and Wei Xue
- 17:10-17:30 Closing