Search NASA⌕ Search

SEARCH · Search NASA

Results for “PROCESSORS”

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 1,495 records · Page 83

Characterizing the impact of using spare-cores on application performance

Increased parallelism on a single processor is driving improvements in peak-performance at both the node and system levels. However achievable performance, in particular from production scientific applications, is not always directly proportional to the core count. Performance is often limited by constraints in the memory hierarchy and also by a node interconnectivity. Even on state-of-the-art processors, containing between four and eight cores, many applications cannot take full advantage of the compute-performance of all cores. This trend is expected to increase on future processors as the core count per processor increases. In this work we characterize the use of spare-cores, cores that do not provide any improvements in application performance, on current multi-core processors. By using a pulse-width modulation method, we examine the possible performance profile of using a spare-core and quantify under what situations its use will not impact application performance. We show that, for current AMD and Intel multi-core processors, spare-cores can be used for substantial computational tasks but can impact application performance when using shared caches or when significantly accessing main memory.

97 MATHEMATICS AND COMPUTING↗

Multi-fault Tolerance for Cartesian Data Distributions

Faults are expected to play an increasingly important role in how algorithms and applications are designed to run on future extreme-scale sys- tems. Algorithm-based fault tolerance (ABFT) is a promising approach that involves modications to the algorithm to recover from faults with lower over- heads than replicated storage and a signicant reduction in lost work compared to checkpoint-restart techniques. Fault-tolerant linear algebra (FTLA) algo- rithms employ additional processors that store parities along the dimensions of a matrix to tolerate multiple, simultaneous faults. Existing approaches as- sume regular data distributions (blocked or block-cyclic) with the failures of each data block being independent. To match the characteristics of failures on parallel computers, we extend these approaches to mapping parity blocks in several important ways. First, we handle parity computation for generalized Cartesian data distributions with each processor holding arbitrary subsets of blocks in a Cartesian-distributed array. Second, techniques to handle corre- lated failures, i.e., multiple processors that can be expected to fail together, are presented. Third, we handle the colocation of parity blocks with the data blocks and do not require them to be on additional processors. Several al- ternative approaches, based on graph matching, are presented that attempt to balance the memory overhead on processors while guaranteeing the same fault tolerance properties as existing approaches that assume independent fail- ures on regular blocked data distributions. The evaluation of these algorithms demonstrates that the additional desirable properties are provided by the pro- posed approach with minimal overhead.

Ali, Nawab↗

Measuring the capabilities of quantum computers

Quantum computers can now run interesting programs, but each processor’s capability—the set of programs that it can run successfully—is limited by hardware errors. These errors can be complicated, making it difficult to accurately predict a processor’s capability. Benchmarks can be used to measure capability directly, but current benchmarks have limited flexibility and scale poorly to many-qubit processors. We show how to construct scalable, efficiently verifiable benchmarks based on any program by using a technique that we call circuit mirroring. With it, we construct two flexible, scalable volumetric benchmarks based on randomized and periodically ordered programs. We use these benchmarks to map out the capabilities of twelve publicly available processors, and to measure the impact of program structure on each one. We find that standard error metrics are poor predictors of whether a program will run successfully on today’s hardware, and that current processors vary widely in their sensitivity to program structure.

97 MATHEMATICS AND COMPUTING↗

Fast and Accurate Simulation of the Cray XMT Multithreaded Supercomputer

Irregular applications, such as data mining and analysis or graph-based computations, show unpredictable memory/network access patterns and control structures. Highly multithreaded architectures with large processor counts, like the Cray MTA-1, MTA-2 and XMT, appear to address their requirements better than commodity clusters. However, the research on highly multithreaded systems is currently limited by the lack of adequate architectural simulation infrastructures due to issues such as size of the machines, memory footprint, simulation speed, accuracy and customization. At the same time, Shared-memory MultiProcessors (SMPs) with multi-core processors have become an attractive platform to simulate large scale machines. In this paper, we introduce a cycle-level simulator of the highly multithreaded Cray XMT supercomputer. The simulator runs unmodified XMT applications. We discuss how we tackled the challenges posed by its development, detailing the techniques introduced to make the simulation as fast as possible while maintaining a high accuracy. By mapping XMT processors (ThreadStorm with 128 hardware threads) to host computing cores, the simulation speed remains constant as the number of simulated processors increases, up to the number of available host cores. The simulator supports zero-overhead switching among different accuracy levels at run-time and includes a network model that takes into account contention. On a modern 48-core SMP host, our infrastructure simulates a large set of irregular applications 500 to 2000 times slower than real time when compared to a 128-processor XMT, while remaining within 10\% of accuracy. Emulation is only from 25 to 200 times slower than real time.

Cray XMT↗

Accurate, Fast and Scalable Kernel Ridge Regression on Parallel and Distributed Systems

Kernel Ridge Regression (KRR) is a fundamental method in machine learning. Given an n-by-d data matrix as input, a traditional implementation requires Θ(n2) memory to form an n-by-n kernel matrix and Θ(n3) flops to compute the final model. These time and storage costs prohibit KRR from scaling up to large datasets. For example, even on a relatively small dataset (a 520k-by-90 input requiring 357 MB), KRR requires 2 TB memory just to store the kernel matrix. Additionally, the reason is that n usually is much larger than d for real-world applications. On the other hand, weak scaling becomes a problem: if we keep d and n/p fixed as p grows (p is # machines), the memory needed grows as Θ(p) per processor and the flops as Θ(p2) per processor. In the perfect weak scaling situation, both the memory needed and the flops grow as Θ(1) per processor (i.e. memory and flops are constant). The traditional Distributed KRR implementation (DKRR) only achieved 0.32% weak scaling efficiency from 96 to 1536 processors. In this work, we propose two new methods to address these problems: the Balanced KRR (BKRR) and K-means KRR (KKRR). These methods consider alternative ways to partition the input dataset into p different parts, generating p different models, and then selecting the best model among them. Compared to a conventional implementation, KKRR2 (optimized version of KKRR) improves the weak scaling efficiency from 0.32% to 38% and achieves a 591x speedup for getting the same accuracy by using the same data and the same hardware (1536 processors). BKRR2 (optimized version of BKRR) achieves a higher accuracy than the current fastest method using less training time for a variety of datasets. For the applications requiring only approximate solutions, BKRR2 improves the weak scaling efficiency to 92% and achieves 3505x speedup (theoretical speedup: 4096x).

97 MATHEMATICS AND COMPUTING↗

Accelerating Scientific Computing in the Post-Moore’s Era

Novel uses of graphical processing units for accelerated computation revolutionized the field of high-performance scientific computing by providing specialized workflows tailored to algorithmic requirements. As the era of Moore’s law draws to a close, many new non–von Neumann processors are emerging as potential computational accelerators, including those based on the principles of neuromorphic computing, tensor algebra, and quantum information. While development of these new processors is continuing to mature, the potential impact on accelerated computing is anticipated to be profound. We discuss how different processing models can advance computing in key scientific paradigms: machine learning and constraint satisfaction. Significantly, each of these new processor types utilizes a fundamentally different model of computation, and this raises questions about how to best use such processors in the design and implementation of applications. While many processors are being developed with a specific domain target, the ubiquity of spin-glass models and neural networks provides an avenue for multi-functional applications. Furthermore, this also hints at the infrastructure needed to integrate next-generation processing units into future high-performance computing systems.

97 MATHEMATICS AND COMPUTING↗

EAP Patterns

EAP Patterns: Memory access and iteration patterns from the EAP code base with the physics removed. This is intended to be a serial app representing memory access patterns. We will populate the data structures from EAP output files to provide representative patterns of face and cell loops within the EAP code base The arguments to the program are the EAP output file and the number of MPI processors to emulate. Currently there is no MPI in the application and the number of processors represents a means of emulating the halo (clone) cells around the domain of the given processor. An optional argument `processor_ID` can be provided to specify the processor to emulate. In the future we plan to include the ability to run multiple MPI ranks to more accurately represent the on-node demands on memory bandwidth.

Swaminarayan, Sriram↗

A look at the performance of AMD Shanghai

This document summarizes a performance comparison between a four processor Shanghai node, a four-processor Barcelona node, and a two processor Nehalem node. All of the processors contain four cores. A comparison of achievable memory bandwidths, and application performance on a single-core, a single-processor (4-cores), and a single node are detailed.

97 MATHEMATICS AND COMPUTING↗

An Early Quantum Computing Proposal

The D-Wave 2X is the third generation of quantum processing created by D-Wave. NASA (with Google and USRA) and Lockheed Martin (with USC), both own D-Wave systems. Los Alamos National Laboratory (LANL) purchased a D-Wave 2X in November 2015. The D-Wave 2X processor contains (nominally) 1152 quantum bits (or qubits) and is designed to specifically perform quantum annealing, which is a well-known method for finding a global minimum of an optimization problem. This methodology is based on direct execution of a quantum evolution in experimental quantum hardware. While this can be a powerful method for solving particular kinds of problems, it also means that the D-Wave 2X processor is not a general computing processor and cannot be programmed to perform a wide variety of tasks. It is a highly specialized processor, well beyond what NNSA currently thinks of as an “advanced architecture.”A D-Wave is best described as a quantum optimizer. That is, it uses quantum superposition to find the lowest energy state of a system by repeated doses of power and settling stages. The D-Wave produces multiple solutions to any suitably formulated problem, one of which is the lowest energy state solution (global minimum). Mapping problems onto the D-Wave requires defining an objective function to be minimized and then encoding that function in the Hamiltonian of the D-Wave system. The quantum annealing method is then used to find the lowest energy configuration of the Hamiltonian using the current D-Wave Two, two-level, quantum processor. This is not always an easy thing to do, and the D-Wave Two has significant limitations that restrict problem sizes that can be run and algorithmic choices that can be made. Furthermore, as more people are exploring this technology, it has become clear that it is very difficult to come up with general approaches to optimization that can both utilize the D-Wave and that can do better than highly developed algorithms on conventional computers for specific applications. These are all fundamental challenges that must be overcome for the D-Wave, or similar, quantum computing technology to be broadly applicable.

97 MATHEMATICS AND COMPUTING↗

CTF Preprocessor User's Manual

This document describes how a user should go about using the CTF pre- processor tool to create an input deck for modeling rod-bundle geometry in CTF. The tool was designed to generate input decks in a quick and less error-prone manner for CTF. The pre-processor is a completely independent utility, written in Fortran, that takes a reduced amount of input from the user. The information that the user must supply is basic information on bundle geometry, such as rod pitch, clad thickness, and axial location of spacer grids--the pre-processor takes this basic information and determines channel placement and connection information to be written to the input deck, which is the most time-consuming and error-prone segment of creating a deck. Creation of the model is also more intuitive, as the user can specify assembly and water-tube placement using visual maps instead of having to place them by determining channel/channel and rod/channel connections. As an example of the benefit of the pre-processor, a quarter-core model that contains 500,000 scalar-mesh cells was read into CTF from an input deck containing 200,000 lines of data. This 200,000 line input deck was produced automatically from a set of pre-processor decks that contained only 300 lines of data.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Effective Vectorization with OpenMP 4.5

This paper describes how the Single Instruction Multiple Data (SIMD) model and its extensions in OpenMP work, and how these are implemented in different compilers. Modern processors are highly parallel computational machines which often include multiple processors capable of executing several instructions in parallel. Understanding SIMD and executing instructions in parallel allows the processor to achieve higher performance without increasing the power required to run it. SIMD instructions can significantly reduce the runtime of code by executing a single operation on large groups of data. The SIMD model is so integral to the processor s potential performance that, if SIMD is not utilized, less than half of the processor is ever actually used. Unfortunately, using SIMD instructions is a challenge in higher level languages because most programming languages do not have a way to describe them. Most compilers are capable of vectorizing code by using the SIMD instructions, but there are many code features important for SIMD vectorization that the compiler cannot determine at compile time. OpenMP attempts to solve this by extending the C++/C and Fortran programming languages with compiler directives that express SIMD parallelism. OpenMP is used to pass hints to the compiler about the code to be executed in SIMD. This is a key resource for making optimized code, but it does not change whether or not the code can use SIMD operations. However, in many cases critical functions are limited by a poor understanding of how SIMD instructions are actually implemented, as SIMD can be implemented through vector instructions or simultaneous multi-threading (SMT). We have found that it is often the case that code cannot be vectorized, or is vectorized poorly, because the programmer does not have sufficient knowledge of how SIMD instructions work.

97 MATHEMATICS AND COMPUTING↗

PLATO (Parallel Load Assignment Tool): A Parallel Workload Partitioner for Particle-based Methods

A key first step in many large-scale high-performance scientific computing applications is the distribution of the computational domain across all participating processors of a parallel computing platform such as a shared memory machine, cluster or a tightly-coupled supercomputer. The primary goal of such a partitioning step is to ensure that the computational work performed by each process remains load-balanced throughout the execution of the parallel application. In this project, a general-purpose parallel partitioning tool, code-named PLATO, was designed and implemented. PLATO assumes that the workload is a set of points in an n-dimensional metric space. These points are abstractions of computational units that may represent point-like computational units such as celestial objects in astrophysics, atoms in materials sciences and molecules in biology/chemistry, or may represent more abstract computational units such as grid points in finite difference methods, grid elements in finite element methods or n-dimensional points in the feature space of a host of other applications. PLATO supports several key and useful capabilities, namely: (a) multiple load partitioning strategies, (b) support for arbitrary user-defined processor topology, (c) pre-fetching data from neighboring processors based on user-defined parameters, and (d) support for arbitrary number and types of attributes associated with each point in the computational domain. PLATO is implemented in the C programming language using Message Passing Interface (MPI) for inter-processor communications and its parallel performance tested on up to 48 processors of a shared memory machine. This report includes a detailed performance study of PLATO using parallel speedup and iso-efficiency as the two main metrics of parallel performance. In addition, a user manual and DOXYGEN generated code documentation have been included for completeness.

97 MATHEMATICS AND COMPUTING↗

CTF User's Manual (V.4.2)

This document describes how a user should go about using the CTF pre-processor tool to create an input deck for modeling rod-bundle geometry in CTF. The tool was designed to generate input decks in a quick and less error-prone manner for CTF. The pre-processor is a completely independent utility, written in Fortran, that takes a reduced amount of input from the user. The information that the user must supply is basic information on bundle geometry, such as rod pitch, clad thickness, and axial location of spacer grids - the pre-processor takes this basic information and determines channel placement and connection information to be written to the input deck, which is the most time-consuming and error-prone segment of creating a deck. Creation of the model is also more intuitive, as the user can specify assembly and water-tube placement using visual maps instead of having to place them by determining channel/channel and rod/channel connections. As an example of the benefit of the pre-processor, a quarter-core model that contains 500,000 scalar-mesh cells was read into CTF from an input deck containing 200,000 lines of data. This 200,000 line input deck was produced automatically from a set of pre-processor decks that contained only 300 lines of data.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

LDRD final report : massive multithreading applied to national infrastructure and informatics.

Large relational datasets such as national-scale social networks and power grids present different computational challenges than do physical simulations. Sandia's distributed-memory supercomputers are well suited for solving problems concerning the latter, but not the former. The reason is that problems such as pattern recognition and knowledge discovery on large networks are dominated by memory latency and not by computation. Furthermore, most memory requests in these applications are very small, and when the datasets are large, most requests miss the cache. The result is extremely low utilization. We are unlikely to be able to grow out of this problem with conventional architectures. As the power density of microprocessors has approached that of a nuclear reactor in the past two years, we have seen a leveling of Moores Law. Building larger and larger microprocessor-based supercomputers is not a solution for informatics and network infrastructure problems since the additional processors are utilized to only a tiny fraction of their capacity. An alternative solution is to use the paradigm of massive multithreading with a large shared memory. There is only one instance of this paradigm today: the Cray MTA-2. The proposal team has unique experience with and access to this machine. The XMT, which is now being delivered, is a Red Storm machine with up to 8192 multithreaded 'Threadstorm' processors and 128 TB of shared memory. For many years, the XMT will be the only way to address very large graph problems efficiently, and future generations of supercomputers will include multithreaded processors. Roughly 10 MTA processor can process a simple short paths problem in the time taken by the Gordon Bell Prize-nominated distributed memory code on 32,000 processors of Blue Gene/Light. We have developed algorithms and open-source software for the XMT, and have modified that software to run some of these algorithms on other multithreaded platforms such as the Sun Niagara and Opteron multi-core chips.

97 MATHEMATICS AND COMPUTING↗

Pulse-level noisy quantum circuits with QuTiP

The study of the impact of noise on quantum circuits is especially relevant to guide the progress of Noisy Intermediate-Scale Quantum (NISQ) computing. In this paper, we address the pulse-level simulation of noisy quantum circuits with the Quantum Toolbox in Python (QuTiP). We introduce new tools in qutip-qip, QuTiP's quantum information processing package. These tools simulate quantum circuits at the pulse level, leveraging QuTiP's quantum dynamics solvers and control optimization features. We show how quantum circuits can be compiled on simulated processors, with control pulses acting on a target Hamiltonian that describes the unitary evolution of the physical qubits. Various types of noise can be introduced based on the physical model, e.g., by simulating the Lindblad density-matrix dynamics or Monte Carlo quantum trajectories. In particular, the user can define environment-induced decoherence at the processor level and include noise simulation at the level of control pulses. We illustrate how the Deutsch-Jozsa algorithm is compiled and executed on a superconducting-qubit-based processor, on a spin-chain-based processor and using control optimization algorithms. We also show how to easily reproduce experimental results on cross-talk noise in an ion-based processor, and how a Ramsey experiment can be modeled with Lindblad dynamics. Finally, we illustrate how to integrate these features with other software frameworks.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Optimizing Performance of Combustion Chemistry Solvers on Intel's Many Integrated Core (MIC) Architectures

This work investigates novel algorithm designs and optimization techniques for restructuring chemistry integrators in zero and multidimensional combustion solvers, which can then be effectively used on the emerging generation of Intel's Many Integrated Core/Xeon Phi processors. These processors offer increased computing performance via large number of lightweight cores at relatively lower clock speeds compared to traditional processors (e.g. Intel Sandybridge/Ivybridge) used in current supercomputers. This style of processor can be productively used for chemistry integrators that form a costly part of computational combustion codes, in spite of their relatively lower clock speeds. Performance commensurate with traditional processors is achieved here through the combination of careful memory layout, exposing multiple levels of fine grain parallelism and through extensive use of vendor supported libraries (Cilk Plus and Math Kernel Libraries). Important optimization techniques for efficient memory usage and vectorization have been identified and quantified. These optimizations resulted in a factor of ~ 3 speed-up using Intel 2013 compiler and ~ 1.5 using Intel 2017 compiler for large chemical mechanisms compared to the unoptimized version on the Intel Xeon Phi. The strategies, especially with respect to memory usage and vectorization, should also be beneficial for general purpose computational fluid dynamics codes.

97 MATHEMATICS AND COMPUTING↗

System support software for the Space Ultrareliable Modular Computer (SUMC)

The highly transportable programming system designed and implemented to support the development of software for the Space Ultrareliable Modular Computer (SUMC) is described. The SUMC system support software consists of program modules called processors. The initial set of processors consists of the supervisor, the general purpose assembler for SUMC instruction and microcode input, linkage editors, an instruction level simulator, a microcode grid print processor, and user oriented utility programs. A FORTRAN 4 compiler is undergoing development. The design facilitates the addition of new processors with a minimum effort and provides the user quasi host independence on the ground based operational software development computer. Additional capability is provided to accommodate variations in the SUMC architecture without consequent major modifications in the initial processors.

Hill, T. E.↗

On-board image correction

This paper describes a processor architecture for performing on-board geometric and radiometric correction of Landsat imagery. The design uses a general purpose processor to calculate the distortion values at selected points in the image and a special purpose processor to resample (calculate distortion at each image point and interpolate the intensity) the sensor output data. A distinct special purpose processor is used for each spectral band. Because of the sensor's high output data rate, 80 M bit per second, the special purpose processors use a pipeline architecture. Sizing has been done of both the general and special purpose hardware.

Martin, D. R.↗