Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “fault tolerant computing”

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 685 records · Page 38

Unitary Qubit Lattice Algorithms for Plasma Physics

This final technical report summarizes research conducted under DOE Award DE-SC0021653 to develop unitary Quantum Lattice Algorithms for modeling electromagnetic wave propagation and scattering in complex media, including plasmas. The project developed and validated quantum-inspired formulations of Maxwell's equations that preserve unitary evolution and can be evaluated on classical high-performance computing systems while providing a foundation for future quantum-computing implementations. Major accomplishments include the development of two- and three-dimensional algorithms for electromagnetic scattering; scalable, distributed-memory implementations demonstrated on the Perlmutter supercomputer; formulations for nonlinear lossless fluid dynamics and cold, lossless, inhomogeneous magnetized plasmas; and an explicit quantum algorithm for a time-discretized Lorenz model. Simulations reproduced a range of characteristic wave phenomena, including transient effects that are not readily apparent in conventional frequency-domain studies, demonstrating the effectiveness of the proposed approach for modeling complex electromagnetic and plasma systems. The work establishes a unified theoretical and computational framework for quantum and quantum-inspired simulation and provides a foundation for future implementation on fault-tolerant quantum systems.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Performance and power modeling and prediction using MuMMI and 10 machine learning methods

Energy-efficient scientific applications require insight into how high performance computing system features impact the applications' power and performance. This insight can result from the development of performance and power models. Here, in this article, we use the modeling and prediction tool MuMMI (Multiple Metrics Modeling Infrastructure) and 10 machine learning methods to model and predict performance and power consumption and compare their prediction error rates. We use an algorithm-based fault-tolerant linear algebra code and a multilevel checkpointing fault-tolerant heat distribution code to conduct our modeling and prediction study on the Cray XC40 Theta and IBM BG/Q Mira at Argonne National Laboratory and the Intel Haswell cluster Shepard at Sandia National Laboratories. Our experimental results show that the prediction error rates in performance and power using MuMMI are less than 10% for most cases. By utilizing the models for runtime, node power, CPU power, and memory power, we identify the most significant performance counters for potential application optimizations, and we predict theoretical outcomes of the optimizations. Based on two collected datasets, we analyze and compare the prediction accuracy in performance and power consumption using MuMMI and 10 machine learning methods.

97 MATHEMATICS AND COMPUTING↗

Distributed execution of recovery blocks - An approach for uniform treatment of hardware and software faults in real-time applications

The concept of distributed execution of recovery blocks is examined as an approach for uniform treatment of hardware and software faults. A useful characteristic of the approach is the relatively small time cost it requires. The approach is thus suitable for incorporation into real-time computer systems. A specific formulation of the approach that is aimed at minimizing the recovery time is presented, called the distributed recovery block (DRB) scheme. The DRB scheme is capable of effecting forward recovery while handling both hardware and software faults in a uniform manner. An approach to incorporating the capability for multiprocessing scheme is also discussed. Two experiments aimed at testing the execution efficiency of the scheme in real-time applications have been conducted on two different multimicrocomputer networks. The results clearly indicate the feasibility of achieving tolerance of hardware and software faults in a broad range of real-time computer systems by use of the schemes for distributed execution of recovery blocks.

Kim, K. H.↗

Micro-Inspector Avionics Module (MAM): A Self-Contained Low Power, Reconfigurable Avionics Platform for Small Spacecrafts and Instruments

This paper describes development of a radiation tolerant, low power, reconfigurable avionics module aimed at meeting the avionics needs of the JPL Micro-Inspector spacecraft. This module represents a complete avionics system, consisting of two PowerPC 405 CPUs embedded within a reconfigurable FPGA fabric of over 8 Million logic gates, 64MB of EDAC protected Flash storage and 128MB of EDAC protected DDR SDRAM or SDRAM memories, along with FPGA SEU mitigation logic, and all necessary power conversion. Processor SEU mitigation is achieved by running the two processors in a lock-step and compare configuration. All of these building blocks are integrated into a double sided circuit board that takes as little as 6 square inches of board space. This module can be embedded into a user system as part of a bigger circuit assembly or as a self contained module. This module is being developed as part of a JPL led Micro-Inspector Program, funded by NASA ESMD aimed at producing a 10Kg micro spacecraft.

configurable↗

Evaluation of fault-tolerant parallel-processor architectures over long space missions

The impact of a five year space mission environment on fault-tolerant parallel processor architectures is examined. The target application is a Strategic Defense Initiative (SDI) satellite requiring 256 parallel processors to provide the computation throughput. The reliability requirements are that the system still be operational after five years with .99 probability and that the probability of system failure during one-half hour of full operation be less than 10(-7). The fault tolerance features an architecture must possess to meet these reliability requirements are presented, many potential architectures are briefly evaluated, and one candidate architecture, the Charles Stark Draper Laboratory's Fault-Tolerant Parallel Processor (FTPP) is evaluated in detail. A methodology for designing a preliminary system configuration to meet the reliability and performance requirements of the mission is then presented and demonstrated by designing an FTPP configuration.

Johnson, Sally C.↗

Performance analysis and optimization for scalable deployment of deep learning models for country-scale settlement mapping on Titan supercomputer

Here, we present a scalable object detection workflow for detecting objects, such as settlements, from remotely sensed (RS) imagery. We have successfully deployed this workflow on Titan supercomputer and utilized it for the task of mapping human settlement at a country scale. The performance of various stages in the workflow was analyzed before making it operational. The workflow implemented various strategies to address issues such as suboptimal resource utilization and long-tail effects due to unbalanced image workload, data loss due to runtime failures, and maximum wall-time constraints imposed by Titan's job scheduling policy. A mean shift clustering–based static load balancing strategy was implemented, which partitions the image load such that each partition contained similar-sized images. Furthermore, a checkpoint-restart strategy was added in the workflow as a fault-tolerance mechanism to prevent the data losses due to unforeseen runtime failures. The performance of the above-mentioned strategies was observed in various scenarios, such as node failure, exceeding wall time, and successful completion. Using this workflow, we have examined an RS data set that has a spatial resolution of 0.31 m and is comprised of 685 675 km 2 of area of the Republic of Zambia in under six hours using 5426 nodes of the Titan supercomputer.

97 MATHEMATICS AND COMPUTING↗

The theory of diversity and redundancy in information system security : LDRD final report.

The goal of this research was to explore first principles associated with mixing of diverse implementations in a redundant fashion to increase the security and/or reliability of information systems. Inspired by basic results in computer science on the undecidable behavior of programs and by previous work on fault tolerance in hardware and software, we have investigated the problem and solution space for addressing potentially unknown and unknowable vulnerabilities via ensembles of implementations. We have obtained theoretical results on the degree of security and reliability benefits from particular diverse system designs, and mapped promising approaches for generating and measuring diversity. We have also empirically studied some vulnerabilities in common implementations of the Linux operating system and demonstrated the potential for diversity to mitigate these vulnerabilities. Our results provide foundational insights for further research on diversity and redundancy approaches for information systems.

99 GENERAL AND MISCELLANEOUS↗

Towards Precision-Aware Fault Tolerance Approaches for Mixed-Precision Applications

Graphics Processing Units (GPUs), the dominantly adopted accelerators in HPC systems, are susceptible to transient hardware fault. New generation of GPUs feature mixed-precision architectures such as NVIDIA Tensor Cores to accelerate matrix multiplications. While being widely adapted, how would they behave under transient hardware faults remain unclear. In this study, we conduct a large-scale fault injection experiments on GEMM kernels implemented with different floating-point data types on the V100 and A100 Tensor Cores, and show distinct error resilience characteristics for the GEMMS with different formats. In the future, we plan to explore this space by building precision-aware floating-point fault tolerance techniques for applications such as DNNs that exercise low-precision computations.

Fang, Bo↗

A High Performance Computing Network and System Simulator for the Power Grid: NGNS^2

Designing and planing next generation power grid sys- tems composed of large power distribution networks, monitoring and control networks, autonomous generators and consumers of power requires advanced simulation infrastructures. The objective is to predict and analyze in time the behavior of networks of systems for unexpected events such as loss of connectivity, malicious attacks and power loss scenarios. This ultimately allows one to answer questions such as: “What could happen to the power grid if ...”. We want to be able to answer as many questions as possible in the shortest possible time for the largest possible systems. In this paper we present a new High Performance Computing (HPC) oriented simulation infrastructure named Next Generation Network and System Simulator (NGNS2 ). NGNS2 allows for the distribution of a single simulation among multiple computing elements by using MPI and OpenMP threads. NGNS2 provides extensive configuration, fault tolerant and load balancing capabilities needed to simulate large and dynamic systems for long periods of time. We show the preliminary results of the simulator running approximately two million simulated entities both on a 64-node commodity Infiniband cluster and a 48-core SMP workstation.

Simulation, power grid, netowrks↗

Resource-Optimized Fermionic Local-Hamiltonian Simulation on a Quantum Computer for Quantum Chemistry

The ability to simulate a fermionic system on a quantum computer is expected to revolutionize chemical engineering, materials design, nuclear physics, to name a few. Thus, optimizing the simulation circuits is of significance in harnessing the power of quantum computers. Here, we address this problem in two aspects. In the fault-tolerant regime, we optimize the R z and T gate counts along with the ancilla qubit counts required, assuming the use of a product-formula algorithm for implementation. We obtain a savings ratio of two in the gate counts and a savings ratio of eleven in the number of ancilla qubits required over the state of the art. In the pre-fault tolerant regime, we optimize the two-qubit gate counts, assuming the use of the variational quantum eigensolver (VQE) approach. Specific to the latter, we present a framework that enables bootstrapping the VQE progression towards the convergence of the ground-state energy of the fermionic system. This framework, based on perturbation theory, is capable of improving the energy estimate at each cycle of the VQE progression, by about a factor of three closer to the known ground-state energy compared to the standard VQE approach in the test-bed, classically-accessible system of the water molecule. The improved energy estimate in turn results in a commensurate level of savings of quantum resources, such as the number of qubits and quantum gates, required to be within a pre-specified tolerance from the known ground-state energy. We also explore a suite of generalized transformations of fermion to qubit operators and show that resource-requirement savings of up to more than 20 % , in small instances, is possible.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Automation of reliability evaluation procedures through CARE - The computer-aided reliability estimation program.

Description of an on-line interactive computer program called CARE (Computer-Aided Reliability Estimation) which can model self-repair and fault-tolerant organizations and perform certain other functions. Essentially CARE consists of a repository of mathematical equations defining the various basic redundancy schemes. These equations, under program control, are then interrelated to generate the desired mathematical model to fit the architecture of the system under evaluation. The mathematical model is then supplied with ground instances of its variables and is then evaluated to generate values for the reliability-theoretic functions applied to the model.

Mathur, F. P.↗

Space Tug Avionics Definition Study. Volume 5: Cost and Programmatics

The baseline avionics system features a central digital computer that integrates the functions of all the space tug subsystems by means of a redundant digital data bus. The central computer consists of dual central processor units, dual input/output processors, and a fault tolerant memory, utilizing internal redundancy and error checking. Three electronically steerable phased arrays provide downlink transmission from any tug attitude directly to ground or via TDRS. Six laser gyros and six accelerometers in a dodecahedron configuration make up the inertial measurement unit. Both a scanning laser radar and a TV system, employing strobe lamps, are required as acquisition and docking sensors. Primary dc power at a nominal 28 volts is supplied from dual lightweight, thermally integrated fuel cells which operate from propellant grade reactants out of the main tanks.

Source record↗

A Conceptual Design for a Reliable Optical Bus (ROBUS)

The Scalable Processor-Independent Design for Electromagnetic Resilience (SPIDER) is a new family of fault-tolerant architectures under development at NASA Langley Research Center (LaRC). The SPIDER is a general-purpose computational platform suitable for use in ultra-reliable embedded control applications. The design scales from a small configuration supporting a single aircraft function to a large distributed configuration capable of supporting several functions simultaneously. SPIDER consists of a collection of simplex processing elements communicating via a Reliable Optical Bus (ROBUS). The ROBUS is an ultra-reliable, time-division multiple access broadcast bus with strictly enforced write access (no babbling idiots) providing basic fault-tolerant services using formally verified fault-tolerance protocols including Interactive Consistency (Byzantine Agreement), Internal Clock Synchronization, and Distributed Diagnosis. The conceptual design of the ROBUS is presented in this paper including requirements, topology, protocols, and the block-level design. Verification activities, including the use of formal methods, are also discussed.

Miner, Paul S.↗

Fault-tolerant parallel processors for avionics with reduced maintenance

Architectural issues and approaches for implementing fault-tolerant parallel processors in avionics systems and other dedicated applications requiring high levels of dependability are discussed. For these systems it is desirable to provide both a high degree of fault tolerance and a system that can operate for extended periods of time without external maintenance. With current technology it is possible to operate dependably for months between maintenance events. The longer-term goal should be maintenance-free operation (a computer that outlasts its host system). It is argued that new high-density packaging techniques may make this type of long-life fault-tolerant design an attractive choice in controlling the life-cycle costs of the next generation of systems.

Rennels, David A.↗

SciDAC - Center for Simulation of Wave Interactions with MHD -- General Atomics Support of ORNL Collaboration

The Center for Simulation of Wave Interactions with Magnetohydrodynamics (SWIM) project is dedicated to conduct research on integrated multi-physics simulations. The Integrated Plasma Simulator (IPS) is a framework that was created by the SWIM team. It provides an integration infrastructure for loosely coupled component-based simulations by facilitating services for code execution coordination, computational resource management, data management, and inter-component communication. The IPS framework features improving resource utilization, implementing application-level fault tolerance, and support of the concurrent multi-tasking execution model. The General Atomics (GA) team worked closely with other team members on this contract, and conducted research in the areas of computational code monitoring, meta-data management, interactive visualization, and user interfaces. The original website to monitor SWIM activity was developed in the beginning of the project. Due to the amended requirements, the software was redesigned and a revision of the website was deployed into production in April of 2010. Throughout the duration of this project, the SWIM Monitoring Portal (http://swim.gat.com:8080/) has been a critical production tool for supporting the project's physics goals.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Performance and scalability evaluation of the Ceph parallel file system. In: PDSW '13 Proceedings of the 8th Parallel Data Storage Workshop

Ceph is an emerging open-source parallel distributed file and storage system. By design, Ceph leverages unreliable commodity storage and network hardware, and provides reliability and fault-tolerance via controlled object placement and data replication. This paper presents our file and block I/O performance and scalability evaluation of Ceph for scientific high-performance computing (HPC) environments. Our work makes two unique contributions. First, our evaluation is performed under a realistic setup for a large-scale capability HPC environment using a commercial high-end storage system. Second, our path of investigation, tuning efforts, and findings made direct contributions to Ceph's development and improved code quality, scalability, and performance. These changes should benefit both Ceph and the HPC community at large.

Wang, Feiyi↗

Demonstration of Algorithmic Quantum Speedup

Despite the development of increasingly capable quantum computers, an experimental demonstration of a provable algorithmic quantum speedup employing today’s non-fault-tolerant devices has remained elusive. Here, in this study, we unequivocally demonstrate such a speedup within the oracular model, quantified in terms of the scaling with the problem size of the time-to-solution metric. We implement the single-shot Bernstein-Vazirani algorithm, which solves the problem of identifying a hidden bitstring that changes after every oracle query, using two different 27-qubit IBM Quantum superconducting processors. The speedup is observed on only one of the two processors when the quantum computation is protected by dynamical decoupling but not without it. The quantum speedup reported here does not rely on any additional assumptions or complexity-theoretic conjectures and solves a bona fide computational problem in the setting of a game with an oracle and a verifier.

97 MATHEMATICS AND COMPUTING↗