In this work, we developed a self-organizing map (SOM) technique for using web-based text analysis to forecast when a group is undergoing a phase change. By 'phase change', we mean that an organization has fundamentally shifted attitudes or behaviors. For instance, when ice melts into water, the characteristics of the substance change. A formerly peaceful group may suddenly adopt violence, or a violent organization may unexpectedly agree to a ceasefire. SOM techniques were used to analyze text obtained from organization postings on the world-wide web. Results suggest it may be possible to forecast phase changes, and determine if an example of writing can be attributed to a group of interest.
In order to provide large quantities of high-reliability disk-based storage, it has become necessary to aggregate disks into fault-tolerant groups based on the RAID methodology. Most RAID levels do provide some fault tolerance, but there are certain classes of applications that require increased levels of fault tolerance within an array. Some of these applications include embedded systems in harsh environments that have a low level of serviceability, or uninhabited data centers servicing cloud computing. When describing RAID reliability, the Mean Time To Data Loss (MTTDL) calculations will often assume that the time to replace a failed disk is relatively low, or even negligible compared to rebuild time. For platforms that are in remote areas collecting and processing data, it may be impossible to access the system to perform system maintenance for long periods. A disk may fail early in a platform's life, but not be replaceable for much longer than typical for RAID arrays. Service periods may be scheduled at intervals on the order of months, or the platform may not be serviced until the end of a mission in progress. Further, this platform may be subject to extreme conditions that can accelerate wear and tear on a disk, requiring even more protection from failures. We have created a high parity RAID implementation that uses a Graphics Processing Unit (GPU) to compute more than two blocks of parity information per stripe, allowing extra parity to eliminate or reduce the requirement for rebuilding data between service periods. While this type of controller is highly effective for RAID 6 systems, an important benefit is the ability to incorporate more parity into a RAID storage system. Such RAID levels, as yet unnamed, can tolerate the failure of three or more disks (depending on configuration) without data loss. While this RAID system certainly has applications in embedded systems running applications in the field, similar benefits can be obtained for servers that are engineered for storage density, with less regard for serviceability or maintainability. A storage brick can be designed to have a MTTDL that extends well beyond the useful lifetime of the hardware used, allowing the disk subsystem to require less service throughout the lifetime of a compute resource. This approach is similar to the Xiotech ISE. Such a design can be deliberately placed remotely (without frequent support) in order to provide colocation, or meet cost goals. For workloads where reliability is key, but conditions are sub-optimal for routine serviceability, a high-parity RAID can provide extra reliability in extraordinary situations. For example, for installations requiring very high Mean Time To Repair, the extra parity can eliminate certain problems with maintaining hot spares, increasing overall reliability. Furthermore, in situations where disk reliability is reduced because of harsh conditions, extra parity can guard against early data loss due to lowered Mean Time To Failure. If used through an iSCSI interface with a streaming workload, it is possible to gain all of these benefits without impacting performance.
In a recent acquisition by DOE/NNSA several large capacity computing clusters called TLCC have been installed at the DOE labs: SNL, LANL and LLNL. TLCC architecture with ccNUMA, multi-socket, multi-core nodes, and InfiniBand interconnect, is representative of the trend in HPC architectures. This paper examines application performance on TLCC contrasting them with Red Storm/Cray XT4. TLCC and Red Storm share similar AMD processors and memory DIMMs. Red Storm however has single socket nodes and custom interconnect. Micro-benchmarks and performance analysis tools help understand the causes for the observed performance differences. Control of processor and memory affinity on TLCC with the numactl utility is shown to result in significant performance gains and is essential to attenuate the detrimental impact of OS interference and cache-coherency overhead. While previous studies have investigated impact of affinity control mostly in the context of small SMP systems, the focus of this paper is on highly parallel MPI applications.
There has been a concerted effort since 2007 to establish a dashboard of metrics for the Science, Technology, and Engineering (ST&E) work at Sandia National Laboratories. These metrics are to provide a self assessment mechanism for the ST&E Strategic Management Unit (SMU) to complement external expert review and advice and various internal self assessment processes. The data and analysis will help ST&E Managers plan, implement, and track strategies and work in order to support the critical success factors of nurturing core science and enabling laboratory missions. The purpose of this SAND report is to provide a guide for those who want to understand the ST&E SMU metrics process. This report provides an overview of why the ST&E SMU wants a dashboard of metrics, some background on metrics for ST&E programs from existing literature and past Sandia metrics efforts, a summary of work completed to date, specifics on the portfolio of metrics that have been chosen and the implementation process that has been followed, and plans for the coming year to improve the ST&E SMU metrics process.
This abstract explores the potential advantages of discontinuous Galerkin (DG) methods for the time-domain inversion of media parameters within the earth's interior. In particular, DG methods enable local polynomial refinement to better capture localized geological features within an area of interest while also allowing the use of unstructured meshes that can accurately capture discontinuous material interfaces. This abstract describes our initial findings when using DG methods combined with Runge-Kutta time integration and adjoint-based optimization algorithms for full-waveform inversion. Our initial results suggest that DG methods allow great flexibility in matching the media characteristics (faults, ocean bottom and salt structures) while also providing higher fidelity representations in target regions. Time-domain inversion using discontinuous Galerkin on unstructured meshes and with local polynomial refinement is shown to better capture localized geological features and accurately capture discontinuous-material interfaces. These approaches provide the ability to surgically refine representations in order to improve predicted models for specific geological features. Our future work will entail automated extensions to directly incorporate local refinement and adaptive unstructured meshes within the inversion process.
Importance sampling is an unbiased sampling method used to sample random variables from different densities than originally defined. These importance sampling densities are constructed to pick 'important' values of input random variables to improve the estimation of a statistical response of interest, such as a mean or probability of failure. Conceptually, importance sampling is very attractive: for example one wants to generate more samples in a failure region when estimating failure probabilities. In practice, however, importance sampling can be challenging to implement efficiently, especially in a general framework that will allow solutions for many classes of problems. We are interested in the promises and limitations of importance sampling as applied to computationally expensive finite element simulations which are treated as 'black-box' codes. In this paper, we present a customized importance sampler that is meant to be used after an initial set of Latin Hypercube samples has been taken, to help refine a failure probability estimate. The importance sampling densities are constructed based on kernel density estimators. We examine importance sampling with respect to two main questions: is importance sampling efficient and accurate for situations where we can only afford small numbers of samples? And does importance sampling require the use of surrogate methods to generate a sufficient number of samples so that the importance sampling process does increase the accuracy of the failure probability estimate? We present various case studies to address these questions.