14th Workshop on Resiliency in High Performance Computing (Resilience) in Clusters, Clouds, and Grids
Overview
Resilience is a critical challenge as high performance computing (HPC)
systems continue to increase component counts, individual component
reliability decreases (such as due to shrinking process technology and
near-threshold voltage (NTV) operation), hardware complexity increases (such
as due to heterogeneous computing) and software complexity increases (such
as due to complex data- and workflows, real-time requirements and
integration of artificial intelligence (AI) technologies with traditional
applications).
Correctness and execution efficiency, in spite of faults, errors, and
failures, is essential to ensure the success of the HPC systems, cluster
computing environments, Grid computing infrastructures, and Cloud computing
services. The impact of faults, errors, and failures in such HPC systems can
range from financial losses due to system downtime (sometimes several
tens-of-thousands of Dollars per lost system-hour), to financial losses due
to unnecessary overprovision (acquisition and operating costs), to financial
losses and legal liabilities due to erroneous or delayed output.
The emergence of AI technology opens up new possibilities, but also new
problems. Using AI technology for operational intelligence that enables
resilience in HPC systems and centers is a complex control problem, while
designing resilient AI technology for HPC applications is a difficult
algorithmic problem. Resilience for HPC systems encompasses a wide spectrum
of fundamental and applied research and development, including theoretical
foundations, error/failure and anomaly detection, monitoring and control,
end-to-end data integrity, enabling infrastructure, and resilient
algorithms.
This workshop brings together experts in the community to further research
and development in HPC resilience and to facilitate exchanges across the
computational paradigms of extreme-scale HPC, cluster computing, Grid
computing, and Cloud computing.
Submission Guidelines
Authors are invited to submit papers electronically in English in PDF
format. Submitted manuscripts should be structured as technical papers and
BETWEEN 10 AND 12 PAGES including figures, tables and references, using
Springer's Lecture Notes in Computer Science (LNCS) format at <https://www.springer.com/gp/computer-science/lncs/conference-proceedings-guidelines>.
Papers with less than 10 or more than 12 pages will not be accepted due to
publisher guidelines. Submissions should include abstract, key words and the
e-mail address of the corresponding author. Papers not conforming to these
guidelines may be returned without review. All manuscripts will be reviewed
and will be judged on correctness, originality, technical strength,
significance, quality of presentation, and interest and relevance to the
conference 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 or not
appropriately structured may also not be considered. The proceedings will be
published in Springer's LNCS as post-conference proceedings.
At least one
author of an accepted paper must register for and attend the workshop for
inclusion in the proceedings. Authors may contact the workshop program
chairs for more information.
Topics
Topics of interest include, but are not limited to:
- Theoretical foundations for resilience:
- Metrics and measurement
- Statistics and optimization
- Simulation and emulation
- Formal methods
- Efficiency modeling and uncertainty quantification
- Experience reports
- Error/failure/anomaly detection and reliability/dependability modeling:
- Statistical analyses
- Machine learning and artificial intelligence
- Digital twins
- Data collection and aggregation
- Information visualization
- Monitoring and control for resilience:
- Center, system and application monitoring and control
- Reliability, availability, serviceability and performability
- Tunable fidelity and quality of service
- Automated response and recovery
- Operational intelligence to enable resilience
- End-to-end integrity:
- Fault tolerant design of centers, systems and applications
- Forward migration and verification
- Degraded operation
- Error propagation, failure cascades, and error/failure containment
- Testing and evaluation, including fault/error/failure injection
- Enabling infrastructure for resilience:
- Reliability, availability, serviceability systems
- System software and middleware
- Resilience extensions for programming models
- Tools and frameworks
- Support for resilience in heterogeneous architectures
- Resilient algorithms:
- Algorithmic detection and correction
- Resilient solvers and algorithm-based fault tolerance
- Fault tolerant numerical methods
- Robust iterative algorithms
- Resilient artificial intelligence
Important Dates
- Workshop papers due:
May 7, 2021 May 14, 2021 June 14, 2021 (23:59 AoE) - Workshop author notification: June 30, 2021
- Workshop author registration: July 21, 2021
- Workshop paper (for informal workshop proceedings, due in EasyChair): July 21, 2021
- Workshop date: August 30, 2021
- Workshop camera-ready papers:
September 10, 2021 September 16, 2021 (23:55 CET - firm deadline)
Workshop Chairs
- Stephen L. Scott
Tennessee Tech University, USA
sscott@tntech.edu - Christian Engelmann
Oak Ridge National Laboratory, USA
engelmannc@ornl.gov
Workshop Program Chairs
- Ferrol Aderholdt
Middle Tennessee State University, USA
ferrol.aderholdt@mtsu.edu - Thomas Naughton
Oak Ridge National Laboratory, USA
naughtont@ornl.gov
Workshop Chair Emeritus
- Chokchai (Box) Leangsuksun
Louisiana Tech University, USA
box@latech.edu
Program Committee
- Wesley Bland, Intel Corporation, USA
- Hans-Joachim Bungartz, Technical University of Munich, Germany
- Marc Casas, Barcelona Supercomputer Center, Spain
- Zizhong Chen, University of California at Riverside, USA
- James Elliott, Sandia National Laboratories, USA
- Kurt Ferreira, Sandia National Laboratories, USA
- Saurabh Hukerikar, NVIDIA, USA
- Ignacio Laguna, Lawrence Livermore National Laboratory, USA
- Scott Levy, Sandia National Laboratories, USA
- Rolf Riesen, Intel Corporation, USA
- Yves Robert, ENS Lyon, France
- Thomas Ropars, Universite Grenoble Alpes, France
- Martin Schulz, Technical University of Munich, Germany
- Keita Teranishi, Sandia National Laboratories, USA
Program (Tentative)
- Euro-Par'21 Workshop Schedule
- 13:30 - 15:00 Lisbon (8:30am - 10:00am US Eastern) Session 1 [Chair: Stephen L. Scott]
- 13:30 - 13:45 (08:30am-08:45am ET) Opening: Resilience Workshop Organizers
- 13:45 - 14:35 (08:45am-09:35am ET) Keynote: Christian Engelmann Faults, Errors and Failures in Extreme-Scale Supercomputers Presenter.
- 14:35 - 15:00 (09:35am-10:00am ET) Jorg Keller and Sebastian Litzinger. Energy-Efficient Execution of Streaming Task Graphs with Parallelizable Tasks on Multicore Platforms with Core Failures
- 15:00 - 15:15 Lisbon (10:00am - 10:15am US Eastern) Coffee Break
- 15:15 - 16:50 Lisbon (10:15am - 11:50am US Eastern) Session 2 [Chair: Thomas Naughton]
- 15:15 - 15:40 (10:15am-10:40am ET) Ioannis Vardas, Manolis Ploumidis and Manolis Marazakis. Exploring the impact of node failures on the resource allocation for parallel jobs
- 15:40 - 16:05 (10:40am-11:05am ET) Mohit Kumar and Christian Engelmann. RDPM: An Extensible Tool for Resilience Design Patterns Modeling
- 16:05 - 16:30 (11:05am-11:30am ET) Kurt Ferreira and Scott Levy. Characterizing Memory Failures Using Benford's Law
- 16:30 - 16:50 (11:30am-11:50pm ET) Q&A Discussion (led by Resilience Workshop Organizers) / Closing remarks