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), and software complexity increases. Application correctness and execution efficiency, in spite of frequent faults, errors, and failures, is essential to ensure the success of the extreme-scale HPC systems, cluster computing environments, Grid computing infrastructures, and Cloud computing services.
While a fault (e.g., a bug or stuck bit) is the cause of an error, its manifestation as a state change is considered an error (e.g., a bad value or incorrect execution), and the transition to an incorrect service is observed as a failure (e.g., an application abort or system crash). A failure in a computing system is typically observed through an application abort or a full/partial service or system outage. A detectable correctable error is often transparently handled by hardware, such as a single bit flip in memory that is protected with single-error correction double-error detection (SECDED) error correcting code (ECC). A detectable uncorrectable error (DUE) typically results in a failure, such as multiple bit flips in the same addressable word that escape SECDED ECC correction, but not detection, and ultimately cause an application abort. An undetectable error (UE) may result in silent data corruption (SDC), e.g., an incorrect application output. There are many other types of hardware and software faults, errors, and failures in computing systems.
Resilience for HPC systems encompasses a wide spectrum of fundamental and applied research and development, including theoretical foundations, fault detection and prediction, monitoring and control, end-to-end data integrity, enabling infrastructure, and resilient solvers and algorithm-based fault tolerance. 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.
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 <http://www.springer.com/computer/lncs?SGWID=0-164-6-793341-0>. 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 of interest include, but are not limited to: