Search NASA⌕ Search

SEARCH · Search NASA

Results for “fault correction”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 505 records · Page 28

LADR: low-cost application-level detector for reducing silent output corruptions

Applications running on future high performance computing (HPC) systems are more likely to experience transient faults due to technology scaling trends with respect to higher circuit density, smaller transistor size and near-threshold voltage (NTV) operations. A transient fault could corrupt application state without warning, possibly leading to incorrect application output. Such errors are called silent data corruptions (SDCs). In this paper, we present LADR, a low-cost application-level SDC detector for scientific applications. LADR protects scientific applications from SDCs by watching for data anomalies in their state variables (those of scientific interest). It employs compile-time data-flow analysis to minimize the number of monitored variables, thereby reducing runtime and memory overheads while maintaining a high level of fault coverage with low false positive rates. We evaluated LADR with 4 scientific workloads and results show that LADR achieved < 80% fault coverage with only ∼ 3% runtime overheads and ∼ 1% memory overheads. As compared to prior state-of-the-art anomaly-based detection methods, SDC achieved comparable or improved fault coverage, but reduced runtime overheads by 21% ∼ 75%, and memory overheads by 35% ∼ 55% for the evaluated workloads. We believe that such an approach with low memory and runtime overheads coupled with attractive detection precision makes LADR a viable approach for assuring the correct output from large-scale high performance simulations.

Chen, Chao↗

ATTNChecker: Highly-Optimized Fault Tolerant Attention for Large Language Model Training

Large Language Models (LLMs) have demonstrated remarkable performance in various natural language processing tasks. However, the training of these models is computationally intensive and susceptible to faults, particularly in the attention mechanism, which is a critical component of transformer-based LLMs. In this paper, we investigate the impact of faults on LLM training, focusing on INF, NaN, and near-INF values in the computation results with systematic fault injection experiments. We observe the propagation patterns of these errors, which can trigger non-trainable states in the model and disrupt training, forcing the procedure to load from checkpoints. To mitigate the impact of these faults, we propose ATTNChecker, the first Algorithm-Based Fault Tolerance (ABFT) technique tailored for the attention mechanism in LLMs. ATTNChecker is designed based on fault propagation patterns of LLM and incorporates performance optimization to adapt to both system reliability and model vulnerability while providing lightweight protection for fast LLM training. Evaluations on four LLMs show that ATTNChecker on average incurs on average 7% overhead on training while detecting and correcting all extreme errors. Compared with the state-of-the-art checkpoint/restore approach, ATTNChecker reduces recovery overhead by up to 49×.

Liang, Yuhang [University of Alabama - Birmingham]↗

Extended defects in GaAs/GaAs 1-x Sb x /GaAs (001) heterostructures

The atomic-scale structure of extended defects in GaAs/GaAs 1-x Sb x /GaAs (001) heterostructures has been characterized using aberration-corrected scanning transmission electron microscopy. The defect located at the tensile-strained GaAs(cap)/GaAs 0.34 Sb 0.66 interface had no edge component in the projection plane and is identified as either a dissociated screw dislocation or a partial dislocation dipole. The associated intrinsic stacking fault is bounded by two 30° Shockley partial dislocations of opposite sign. Another defect, located at the compressively-strained GaAs 0.91 Sb 0.09 /GaAs(substrate) interface, is identified as a dissociated 90° dislocation. The associated intrinsic stacking fault is bounded by a 30° Shockley partial dislocation and a partial dislocation with a Burgers vector of either or , where a is the lattice constant. Unpaired atomic columns observed at the cores of the 30° Shockley partial dislocations indicate that both defects belong primarily to the glide set.

36 MATERIALS SCIENCE↗

Modeling fault diagnosis as the activation and use of a frame system

Twenty pilots with instrument flight ratings were asked to perform a fault-diagnosis task for which they had relevant domain knowledge. The pilots were asked to think out loud as they requested and interpreted information. Performances were then modeled as the activation and use of a frame system. Cognitive biases, memory distortions and losses, and failures to correctly diagnose the problem were studied in the context of this frame system model.

Smith, Philip J.↗

Non-abelian fractional quantum hall effect for fault-resistant topological quantum computation.

Topological quantum computation (TQC) has emerged as one of the most promising approaches to quantum computation. Under this approach, the topological properties of a non-Abelian quantum system, which are insensitive to local perturbations, are utilized to process and transport quantum information. The encoded information can be protected and rendered immune from nearly all environmental decoherence processes without additional error-correction. It is believed that the low energy excitations of the so-called =5/2 fractional quantum Hall (FQH) state may obey non-Abelian statistics. Our goal is to explore this novel FQH state and to understand and create a scientific foundation of this quantum matter state for the emerging TQC technology. We present in this report the results from a coherent study that focused on obtaining a knowledge base of the physics that underpins TQC. We first present the results of bulk transport properties, including the nature of disorder on the 5/2 state and spin transitions in the second Landau level. We then describe the development and application of edge tunneling techniques to quantify and understand the quasiparticle physics of the 5/2 state.

Pan, Wei↗

Relocation of the 2010-2013 near the north coast of Papua earthquake sequence using Modified Joint Hypocenter Determination (MJHD) method

The accuracy of earthquake hypocenter position is necessary to analyze the tectonic conditions. This study aims to: (1) relocate the mainshock and aftershocks of the large earthquakes in Papua region i.e. June 16, 2010, April 21, 2012 and April 06, 2013 earthquake (2) determine the true fault plane, (3) estimate the area of the fracture, and (4) analyze the advantages and disadvantages of relocation with MJHD method in benefits for tectonic studies. This study used Modified Joint Hypocenter Determination (MJHD) method. Using P arrival phase data reported by the BMKG and openly available from website repogempa.bmkg.go.id, we relocated the mainshock of this large significant earthquake and its aftershocks. Then we identified the prefered fault planes from the candidate fault planes provided by the global CMT catalogue. The position of earthquakes was successfully relocated. The earthquakes mostly were clustered around the mainshock. Earthquakes that not clustered around mainshock are considered to be different mechanism from the mainshock. Relocation results indicate that the mainshock fault plane of June 16, 2010 earthquake is a field with strike 332o, dip 80o and −172o slip, the mainshock fault plane of April 21, 2012 earthquake is a field with strike 82o, dip 84o and 2o slip, the mainshock fault plane of April 06, 2013 earthquake is a field with strike 339o, dip 56o and −137o slip. Fault plane area estimated by cross section graphical method is an area of 2816.0 km2 (June 16, 2010), 906.2 km2 (April 21, 2012) and 1984.3 km2 (April 06, 2013). MJHD method has the advantage that it can calculate a lot of earthquakes simultaneously and has a station correction to account for lateral heterogeneity of the earth. This method successfully provides significant changes to improve the position of the depth of earthquakes that most of the hypocenter depth manually specified as a fixed depth (± 10 km). But this method cannot be sure that the hypocenters derived from the same earthquake mechanism.

58 GEOSCIENCES↗

Performance analysis of a generalized concurrent error detection procedure

A general procedure for error detection in complex systems, called the data block capture and analysis monitoring process, is described and analyzed. It is assumed that, in addition to being exposed to potential external fault sources, a complex system will in general always contain embedded hardware and software fault mechanisms which can cause the system to perform incorrect computations and/or produce incorrect output. Thus, in operation, the system continuously moves back and forth between error and no-error states. These external fault sources or internal fault mechanisms are extremely difficult to detect. The data block capture and analysis monitoring process is concerned with detecting deviations from the normal performance of the system, known as errors, which are symptomatic of fault conditions. The process consists of repeatedly recording a fixed amount of data from a set of predetermined observation lines of the system being monitored (i.e., capturing a block of data) and then analyzing the captured block in an attempt to determine whether the system is functioning correctly. The performances of linear, quadratic, and logarithmic data analysis algorithms are rigorously characterized in terms of the probability of correctly detecting an error, the expectation and variance of the number of false alarms per error, and the expectation and variance of the latency in detection of errors. Insight into the nature of the general problem of error detection is obtained.

Blough, Douglas M.↗

NSTAR Ion Thruster and Breadboard Power Processor Functional Integration Test Results

A 2.3 kW Breadboard Power Processing Unit (BBPPU) was developed as part of the NASA Solar Electric Propulsion Technology Application Readiness (NSTAR) Program. The NSTAR program will deliver an electric propulsion system based on a 30 cm xenon ion thruster to the New Millennium (NM) program for use as the primary propulsion system for the initial NM flight. The final development test for the BBPPU, the Functional Integration Test, was carried out to demonstrate all aspects of BBPPU operation with an Engineering Model Thruster. Test objectives included: (1) demonstration and validation of automated thruster start procedures, (2) demonstration of stable closed loop control of the thruster beam current, (3) successful response and recovery to thruster faults, and (4) successful safing of the system during simulated spacecraft faults. These objectives were met over the specified 80-120 VDC input voltage range and 0.5-2.3 output power capability of the BBPPU. Two minor anomalies were noted in discharge and neutralizer keeper current. These anomalies did not affect the stability of the system and were successfully corrected.

Hamley, John A.↗

The Mars Microprobe Mission: Advanced Micro-Avionics for Exploration Surface

The Mars Microprobe Mission is the second spacecraft developed as part of the New Millennium Program deep space missions. The objective of the Microprobe Project is to demonstrate the applicability of key technologies for future planetary missions by developing two probes for deployment on Mars. The probes are designed with a single stage entry, descent, and landing system and impact the Martian surface at speeds of approximately 200 meters per second. The microprobes are composed of two main sections, a forebody section that penetrates to a depth below the Martian surface of 0.5 to 2 meters, and an aftbody section that remains on the surface. Each probe system consists of a number of advanced technology components developed specifically for this mission. These include a non-erosive aeroshell for entry into. the atmosphere, a set of low temperature batteries to supply probe power, an advanced microcontroller to execute the mission sequence, collect the science data, and react to possible system fault conditions, a telecommunications subsystem implemented on a set of custom integrated circuits, and instruments designed to provide science measurements from above and below the Martian surface. All of the electronic components have been designed and fabricated to withstand the severe impact shock environment and to operate correctly at predicted temperatures below -100 C.

Blue, Randel↗

Fault-tolerance for exascale systems.

Periodic, coordinated, checkpointing to disk is the most prevalent fault tolerance method used in modern large-scale, capability class, high-performance computing (HPC) systems. Previous work has shown that as the system grows in size, the inherent synchronization of coordinated checkpoint/restart (CR) limits application scalability; at large node counts the application spends most of its time checkpointing instead of executing useful work. Furthermore, a single component failure forces an application restart from the last correct checkpoint. Suggested alternatives to coordinated CR include uncoordinated CR with message logging, redundant computation, and RAID-inspired, in-memory distributed checkpointing schemes. Each of these alternatives have differing overheads that are dependent on both the scale and communication characteristics of the application. In this work, using the Structural Simulation Toolkit (SST) simulator, we compare the performance characteristics of each of these resilience methods for a number of HPC application patterns on a number of proposed exascale machines. The result of this work provides valuable guidance on the most efficient resilience methods for exascale systems.

99 GENERAL AND MISCELLANEOUS↗

Feedback Control Strategy for Transient Stability Application

Power systems are subjected to a wide range of disturbances during daily operations. Severe disturbances, such as a loss of a large generator, a three-phase bolted fault on a generator bus, or a loss of a transmission line, can lead to the loss of synchronism of a generator or group of generators. The ability of a power system to maintain synchronism during the few seconds after being subjected to a severe disturbance is known as transient stability. Most of the modern methods of controlling transient stability involve special protection schemes or remedial action schemes. These special protection schemes sense predetermined system conditions and take corrective actions, such as generator tripping or generation re-dispatch, in real time to maintain transient stability. Another method is the use of a real-time feedback control system to modulate the output of an actuator in response to a signal. This paper provides a fundamental evaluation of the use of feedback control strategies to improve transient stability in a power system. An optimal feedback control strategy that modulates the real power injected and absorbed by distributed energy-storage devices is proposed. Its performance is evaluated on a four-machine power system and on a 34-machine reduced-order model of the Western North American Power System. The result shows that the feedback control strategy can increase the critical fault clearing time by 60%, thereby improving the transient stability of the power system.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Cluster-based architecture for fault-tolerant quantum computation

We present a detailed description of an architecture for fault-tolerant quantum computation, which is based on the cluster model of encoded qubits. In this cluster-based architecture, concatenated computation is implemented in a quite different way from the usual circuit-based architecture where physical gates are recursively replaced by logical gates with error-correction gadgets. Instead, some relevant cluster states, say fundamental clusters, are recursively constructed through verification and postselection in advance for the higher-level one-way computation, which namely provides error-precorrection of gate operations. A suitable code such as the Steane seven-qubit code is adopted for transversal operations. This concatenated construction of verified fundamental clusters has a simple transversal structure of logical errors, and achieves a high noise threshold {approx}3% for computation by using appropriate verification procedures. Since the postselection is localized within each fundamental cluster with the help of deterministic bare controlled-Z gates without verification, divergence of resources is restrained, which reconciles postselection with scalability.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Intelligent fault-tolerant controllers

A system with fault tolerant controls is one that can detect, isolate, and estimate failures and perform necessary control reconfiguration based on this new information. Artificial intelligence (AI) is concerned with semantic processing, and it has evolved to include the topics of expert systems and machine learning. This research represents an attempt to apply AI to fault tolerant controls, hence, the name intelligent fault tolerant control (IFTC). A generic solution to the problem is sought, providing a system based on logic in addition to analytical tools, and offering machine learning capabilities. The advantages are that redundant system specific algorithms are no longer needed, that reasonableness is used to quickly choose the correct control strategy, and that the system can adapt to new situations by learning about its effects on system dynamics.

Huang, Chien Y.↗

A New On-Line Diagnosis Protocol for the SPIDER Family of Byzantine Fault Tolerant Architectures

This paper presents the formal verification of a new protocol for online distributed diagnosis for the SPIDER family of architectures. An instance of the Scalable Processor-Independent Design for Electromagnetic Resilience (SPIDER) architecture consists of a collection of processing elements communicating over a Reliable Optical Bus (ROBUS). The ROBUS is a specialized fault-tolerant device that guarantees Interactive Consistency, Distributed Diagnosis (Group Membership), and Synchronization in the presence of a bounded number of physical faults. Formal verification of the original SPIDER diagnosis protocol provided a detailed understanding that led to the discovery of a significantly more efficient protocol. The original protocol was adapted from the formally verified protocol used in the MAFT architecture. It required O(N) message exchanges per defendant to correctly diagnose failures in a system with N nodes. The new protocol achieves the same diagnostic fidelity, but only requires O(1) exchanges per defendant. This paper presents this new diagnosis protocol and a formal proof of its correctness using PVS.

Geser, Alfons↗

ML–Enabled FPGA Framework for Fast Quantum State Discrimination in Mid-Circuit Measurement Regimes

Accurate and low-latency quantum state discrimination is essential for protocols involving mid-circuit measurement (MCM) and conditional feed-forward. In superconducting quantum systems, conventional readout pipelines transfer measurement data to host processors for post-processing, introducing millisecond-scale delays that far exceed qubit coherence times. To overcome this bottleneck, we present an in-situ machine learning (ML) inference engine implemented on an FPGA for real-time quantum state discrimination. Our design performs inference directly on digitized readout signals with 40 ns latency, supports both qubit and qutrit readout, and enables conditional operations without host-side intervention. This capability is critical for MCM and for feedback-driven protocols such as quantum error correction. We validate the system on superconducting transmon hardware, demonstrating robust discrimination fidelity across multiple qubit and qutrit channels. We further demonstrate conditional qutrit logic driven by FPGA-resident classification, highlighting the potential of low-latency ML-on-FPGA control for NISQ applications and scalable fault-tolerant quantum computing.

Vora, Neel [Lawrence Berkeley National Laboratory ↗

Locating Anomalies in Complex Data Sets Using Visualization and Simulation

The research goals are to create a simulation framework that can accept any combination of models written at the gate or behavioral level. The framework provides the ability to fault simulate and create scenarios of experiments using concurrent simulation. In order to meet these goals we have had to fulfill the following requirements. The ability to accept models written in VHDL, Verilog or the C languages. The ability to propagate faults through any model type. The ability to create experiment scenarios efficiently without generating every possible combination of variables. The ability to accept adversity of fault models beyond the single stuck-at model. Major development has been done to develop a parser that can accept models written in various languages. This work has generated considerable attention from other universities and industry for its flexibility and usefulness. The parser uses LEXX and YACC to parse Verilog and C. We have also utilized our industrial partnership with Alternative System's Inc. to import vhdl into our simulator. For multilevel simulation, we needed to modify the simulator architecture to accept models that contained multiple outputs. This enabled us to accept behavioral components. The next major accomplishment was the addition of "functional fault models". Functional fault models change the behavior of a gate or model. For example, a bridging fault can make an OR gate behave like an AND gate. This has applications beyond fault simulation. This modeling flexibility will make the simulator more useful for doing verification and model comparison. For instance, two or more versions of an ALU can be comparatively simulated in a single execution. The results will show where and how the models differed so that the performance and correctness of the models may be evaluated. A considerable amount of time has been dedicated to validating the simulator performance on larger models provided by industry and other universities.

Panetta, Karen↗

Deep Moonquake Focal Mechanisms: Recovery and Implications

A defining characteristic of deep moonquakes is their tendency to occur with tidal periodicity, prompting previous studies to infer that they are related to the buildup and release of tidal stress within the Moon. In studies of tidal forcing, a key constraint is the focal mechanism: the fault parameters describing the type of failure moonquakes represent. The quality of the lunar seismic data and the limited source/receiver geometries of the Apollo seismic network prohibit the determination of deep moonquake fault parameters using first-motion polarities, as is typically done in terrestrial seismology. Without being able to resolve tidal stress onto a known failure plane, we can examine only gross qualities of the tidal stress tensor with respect to moonquake occurrence, so we cannot fully address the role of tidal stress in moonquake generation. We will examine the extent to which shear (S) and compression (P) wave amplitude ratios can constrain moonquake fault geometry by determining whether, for a given cluster, there exists a focal mechanism that can produce a radiation pattern consistent with the amplitudes measured by the Apollo instruments. Amplitudes are read in the ray coordinate frame, directly from seismograms for which the P and S arrivals are clearly identifiable on all long-period channels of the four Apollo stations. We apply an empirical station correction to account for site effects and the differences between P- and S-wave attenuation. Instead of focusing on the best fitting solution only, we formulate the inverse problem using a falsification criterion: all source orientations that do not reproduce the observed SV/P ratios within an error margin derived from the uncertainty of amplitude readings are rejected. All others are accepted as possible solutions. The inversion is carried out using an exhaustive grid search on a regular grid with predefined step size, encompassing all possible combinations of strike, dip and slip. To assess the sensitivity of the inversion for the uncertainty of the lunar interior structure, we carry out repeated inversions with different velocity structures. Our data set consist of a total of 106 events from 25 deep moonquake clusters. The largest contribution of 37 events originates from the most active cluster, A001, while other clusters are represented by 1 to 9 events. Since the definition of a cluster implies that all events share the same source orientation, a comparison of the inversion results of all events from one cluster will reduce ambiguities of the inversion. Once we obtain a suite of fault parameters for a given source, we can attempt to further constrain the focal mechanism with refined analyses of tidal stresses and predictions based on synthetic seismograms.

Knapmeyer, Martin↗

Fault tolerance in an inner-outer solver: A GVR-enabled case study

Resilience is a major challenge for large-scale systems. It is particularly important for iterative linear solvers, since they take much of the time of many scientific applications. We show that single bit flip errors in the Flexible GMRES iterative linear solver can lead to high computational overhead or even failure to converge to the right answer. Informed by these results, we design and evaluate several strategies for fault tolerance in both inner and outer solvers appropriate across a range of error rates. We implement them, extending Trilinos’ solver library with the Global View Resilience (GVR) programming model, which provides multi-stream snapshots, multi-version data structures with portable and rich error checking/recovery. Lastly, experimental results validate correct execution with low performance overhead under varied error conditions.

97 MATHEMATICS AND COMPUTING↗