Search NASA⌕ Search

SEARCH · Search NASA

Results for “finite precision”

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 19 records

A high-precision finite element method for shock-tube calculations

A two-pass explicit scheme is developed in order to exploit some of the capabilities of finite difference modeling (FDM) for finite element modeling (FEM), which offers the opportunity to account for any type of geometry in fluid flow modeling. Features of the first-order upwind and the Lax-Wendroff high precision explicit finite difference algorithms are reviewed. A flux limiter is developed for FEM to serve as an analog for the single limiter function which has been defined for the various FDMs. It is shown that an antidiffusive limiter must be introduced into the weighting function which normally multiplies the time-derivative term in the variational equation. The two-pass scheme which results is demonstrated to be the equivalent of FDMs with five-point support. However, the present scheme is valid only for one-dimensional calculations and linear shape functions for shock tube flow phenomena. Further work is required for its use with nonlinear hyperbolic systems.

Hughes, T. J. R.↗

A GPU Accelerated Mixed‐Precision Finite Difference Informed Random Walker (FDiRW) Solver for Strongly Inhomogeneous Diffusion Problems

In nature, many complex multi‐physics coupling problems exhibit significant diffusivity inhomogeneity, where one process occurs several orders of magnitude faster than others temporally. Simulating rapid diffusion alongside slower processes demands intensive computational resources due to the necessity for small time steps. To address these computational challenges, we have developed an efficient numerical solver named Finite Difference informed Random Walker (FDiRW). In this study, we propose a GPU‐accelerated, mixed‐precision configuration for the FDiRW solver to maximize efficiency through GPU multi‐threaded parallel computation and lower precision computation. Numerical evaluation results reveal that the proposed GPU‐accelerated mixed‐precision FDiRW solver can achieve a 117× speedup over the CPU baseline, while an additional 1.75× speedup is achieved by employing lower precision GPU computation. Notably, for large model sizes, the GPU‐accelerated mixed‐precision FDiRW solver demonstrates strong scaling with the number of nodes used in simulation. When simulating radionuclide absorption processes by porous wasteform particles with a medium‐sized model of 192 × 192 × 192, this approach reduces the total computational time to 10 min, enabling the simulation of larger systems with strongly inhomogeneous diffusivity.

97 MATHEMATICS AND COMPUTING↗

Optimal q-Markov COVER for finite precision implementation

The existing q-Markov COVER realization theory does not take into account the problems of arithmetic errors due to both the quantization of states and coefficients of the reduced order model. All q-Markov COVERs allow some freedom in the choice of parameters. Here, researchers exploit this freedom in the existing theory to optimize the models with respect to these finite wordlength effects.

Williamson, Darrell↗

The consistent comparison problem in N-version software

The authors have identified a difficulty in the implementation of N-version programming. The problem, called the consistent comparison problem, arises for applications in which decisions are based on the results of comparing finite-precision numbers. It is shown that when versions make comparisons involving the results of finite-precision calculations, it is impossible to guarantee the consistency of their results. It is therefore possible that correct versions may arrive at completely different outputs for an application that does not apparently have multiple correct solutions. If this problem is not dealt with explicitly, an N-version system may be unable to reach consensus even when none of its component versions falls.

Brilliant, Susan S.↗

Mixed precision s –step Lanczos and conjugate gradient algorithms

Compared to the classical Lanczos algorithm, the s-step Lanczos variant has the potential to improve performance by asymptotically decreasing the synchronization cost per iteration. However, this comes at a price; despite being mathematically equivalent, the s-step variant may behave quite differently in finite precision, potentially exhibiting greater loss of accuracy and slower convergence relative to the classical algorithm. It has previously been shown that the errors in the s-step version follow the same structure as the errors in the classical algorithm, but are amplified by a factor depending on the square of the condition number of the O(s)-dimensional Krylov bases computed in each outer loop. As the condition number of these s-step bases grows (in some cases very quickly) with s, this limits the s values that can be chosen and thus can limit the attainable performance. In this work, we show that if a select few computations in s-step Lanczos are performed in double the working precision, the error terms then depend only linearly on the conditioning of the s-step bases. This has the potential for drastically improving the numerical behavior of the algorithm with little impact on per-iteration performance. Our numerical experiments demonstrate the improved numerical behavior possible with the mixed precision approach, and also show that this improved behavior extends to mixed precision s-step CG. Here, we present preliminary performance results on NVIDIA V100 GPUs that show that the overhead of extra precision is minimal if one uses precisions implemented in hardware.

97 MATHEMATICS AND COMPUTING↗

Preliminary Theoretical Analysis of Mixed Precision Krylov Subspace Methods (Q3 Report)

The third quarter of the project was spent performing theoretical finite precision analysis of Krylov subspace method variants that use mixed precision. Our focus here is on the Conjugate Gradient (CG) method and the Lanczos method. We have performed an analysis of maximum attainable accuracy for the classical CG method in which 3 precisions are used: a working precision ε, a precision ε IP for the inner product computations, and a precision ε MV for the matrix-vector products. Our results show that performing inner product computations in lower precision does not affect the attainable accuracy. Further, we have performed a complete error analysis of the s-step Lanczos algorithm. In this case, we show that the numerical behavior of the algorithm can be significantly improved by using extra precision in a small part of the computation. We summarize the main theorems in the remainder of the document. Other activities include attending biweekly xSDK meetings. The subsequent quarter will be spent finalizing these results into technical reports and/or manuscripts for submission to journals, as well as identifying opportunities for future work.

97 MATHEMATICS AND COMPUTING↗

Verification of Numerical Algorithms

The following strategy is suggested for specification and proof: (1) Defer the construction of a formal program specification with respect to I/O assertions unit the correctness of the program with respect to an abstract mathematical model of program intent is demonstrated. (2) Prove that an abstract machine (using infinite precision arithmetic) would compute that object exactly. (3) Prove that the computational sequences of arithmetic operations that occur in the abstract machine must be precisely the same at every step as those occurring on an actual machine (with finite precision arithmetic), executing the same program. (4) Use a Verification Conditions VC-generator that knows about the semantics of arithmetic operations to annotate the program with assertions that bound (or in some circumstances estimate) the difference between the actual machine state variables and the corresponding ones of the abstract machine. Construct the formal program specification by combining the verification conditions into theorems about computational error that can be proved with mechanical assistance.

Source record↗

Initial Exploration of the Use of Mixed Precision in Iterative Solvers (B639388 Subcontract Quarter I Report)

The first quarter of the project was primarily spent identifying potential projects at the intersection of finite precision analysis, mixed precision computation, and Krylov subspace methods. We summarize our findings in the remainder of the document. Other activities include attending biweekly xSDK meetings as well as contributing material to the technical report and journal versions of the multiprecision landscape paper. The subsequent quarter will be spent selecting a subset of the described projects to focus on, performing initial numerical experiments to evaluate the potential for the use of mixed precision, and developing initial theoretical analysis.

97 MATHEMATICS AND COMPUTING↗

Understanding the many-body expansion for large systems. I. Precision considerations

Electronic structure methods based on low-order “n-body” expansions are an increasingly popular means to defeat the highly nonlinear scaling of ab initio quantum chemistry calculations, taking advantage of the inherently distributable nature of the numerous subsystem calculations. Here, we examine how the finite precision of these subsystem calculations manifests in applications to large systems, in this case, a sequence of water clusters ranging in size up to (H{sub 2}O){sub 47}. Using two different computer implementations of the n-body expansion, one fully integrated into a quantum chemistry program and the other written as a separate driver routine for the same program, we examine the reproducibility of total binding energies as a function of cluster size. The combinatorial nature of the n-body expansion amplifies subtle differences between the two implementations, especially for n ⩾ 4, leading to total energies that differ by as much as several kcal/mol between two implementations of what is ostensibly the same method. This behavior can be understood based on a propagation-of-errors analysis applied to a closed-form expression for the n-body expansion, which is derived here for the first time. Discrepancies between the two implementations arise primarily from the Coulomb self-energy correction that is required when electrostatic embedding charges are implemented by means of an external driver program. For reliable results in large systems, our analysis suggests that script- or driver-based implementations should read binary output files from an electronic structure program, in full double precision, or better yet be fully integrated in a way that avoids the need to compute the aforementioned self-energy. Moreover, four-body and higher-order expansions may be too sensitive to numerical thresholds to be of practical use in large systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Mixed precision efficiency improvements for multi-physics flow simulations

This report summarizes research activities at Cascade Technologies during February 2020 - November 2020 to support DOE SBIR phase I contract DE-SC0020548. The primary technical objectives of the work are two-fold: a) to establish a theoretical framework to understand the propagation and impact of finite-precision-related errors and b) to quantify the trade-offs between accelerator throughput and solution accuracy through a series of validation tests involving multi-physics flow phenomena.

97 MATHEMATICS AND COMPUTING↗

Variable Precision Computing

This report summarizes the activities and major accomplishments of the Variable Precision Computing Strategic Initiative project. The overarching goal of this project was to initiate and promote a new paradigm in High Performance Computing (HPC) that would fundamentally change how we represent and make use of representations of real numbers in finite precision and that would establish Lawrence Livermore National Laboratory (LLNL) as a leader in precision-related research for HPC. We pursued three integrated and concurrent research thrusts: new, more dynamic data representations to address data motion and capacity limitations; improved mixed precision algorithms to accelerate computation and to leverage new hardware; and new tools to help developers reason about precision and to automate code transformations. Within these thrust areas, we made significant advances in the use of compressed array data types in numerical calculations, in the theoretical justification, in improved performance, and in the prospects of hardware implementation; in the development of metrics to significantly reduce the storage needs of molecular dynamics simulations; in the development and analysis of efficient mixed-precision algorithms for time-dependent partial differential equations and for orthogonal factorization; and in the capability of tools that can analyze sensitivity of computed results to choices in precision and that can make automated code transformations based on such analysis. We established that there are definite opportunities to realize significant improvements in HPC above and beyond the 2x limits of traditional mixed precision using the technologies we have developed. The project produced numerous papers, presentations, and posters in important venues as well as open-source software contributions to the community, and many of the new capabilities have initiated or are being leveraged in ongoing projects.

97 MATHEMATICS AND COMPUTING↗

A hidden analytic structure of the Rabi model

The Rabi model describes the simplest interaction between a cavity mode with a frequency ω{sub c} and a two-level system with a resonance frequency ω{sub 0}. It is shown here that the spectrum of the Rabi model coincides with the support of the discrete Stieltjes integral measure in the orthogonality relations of recently introduced orthogonal polynomials. The exactly solvable limit of the Rabi model corresponding to Δ=ω{sub 0}/(2ω{sub c})=0, which describes a displaced harmonic oscillator, is characterized by the discrete Charlier polynomials in normalized energy ϵ, which are orthogonal on an equidistant lattice. A non-zero value of Δ leads to non-classical discrete orthogonal polynomials ϕ{sub k}(ϵ) and induces a deformation of the underlying equidistant lattice. The results provide a basis for a novel analytic method of solving the Rabi model. The number of ca. 1350 calculable energy levels per parity subspace obtained in double precision (cca 16 digits) by an elementary stepping algorithm is up to two orders of magnitude higher than is possible to obtain by Braak’s solution. Any first n eigenvalues of the Rabi model arranged in increasing order can be determined as zeros of ϕ{sub N}(ϵ) of at least the degree N=n+n{sub t}. The value of n{sub t}>0, which is slowly increasing with n, depends on the required precision. For instance, n{sub t}≃26 for n=1000 and dimensionless interaction constant κ=0.2, if double precision is required. Given that the sequence of the lth zeros x{sub nl}’s of ϕ{sub n}(ϵ)’s defines a monotonically decreasing discrete flow with increasing n, the Rabi model is indistinguishable from an algebraically solvable model in any finite precision. Although we can rigorously prove our results only for dimensionless interaction constant κ<1, numerics and exactly solvable example suggest that the main conclusions remain to be valid also for κ≥1. -- Highlights: •A significantly simplified analytic solution of the Rabi model. •The spectrum is the lattice of discrete orthogonal polynomials. •Up to 1350 levels in double precision can be obtained for a given parity. •Omission of any level can be easily detected.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Issues in the digital implementation of control compensators

Techniques developed for the finite-precision implementation of digital filters were used, adapted, and extended for digital feedback compensators, with particular emphasis on steady state, linear-quadratic-Gaussian compensators. Topics covered include: (1) the linear-quadratic-Gaussian problem; (2) compensator structures; (3) architectural issues: serialism, parallelism, and pipelining; (4) finite wordlength effects: quantization noise, quantizing the coefficients, and limit cycles; and (5) the optimization of structures.

Moroney, P.↗

ZFP: A compressed array representation for numerical computations

HPC trends favor algorithms and implementations that reduce data motion relative to FLOPS. We investigate the use of lossy compressed data arrays in place of traditional IEEE floating point arrays to store the primary data of calculations. Simulation is fundamentally an exercise in controlled approximation, and error introduced by finite-precision arithmetic (or lossy compression) is just one of several sources of error that need to be managed to ensure sufficient accuracy in a computed result. We describe ZFP, a compressed numerical format designed for in-memory storage of multidimensional arrays, and summarize theoretical results that demonstrate that the error of repeated lossy compression can be bounded and controlled. Furthermore, we establish a relationship between grid resolution and compression-induced errors and show that, contrary to conventional floating point, ZFP reduces finite-difference errors with finer grids. We present example calculations that demonstrate data reduction by 4x or more with negligible impact on solution accuracy. Our results further demonstrate several orders-of-magnitude increase in accuracy using ZFP over IEEE floating point and Posits for the same storage budget.

Lindstrom, Peter↗

Optimized FPGA Implementation of Multi-Rate FIR Filters Through Thread Decomposition

Multirate (decimation/interpolation) filters are among the essential signal processing components in spaceborne instruments where Finite Impulse Response (FIR) filters are often used to minimize nonlinear group delay and finite-precision effects. Cascaded (multi-stage) designs of Multi-Rate FIR (MRFIR) filters are further used for large rate change ratio, in order to lower the required throughput while simultaneously achieving comparable or better performance than single-stage designs. Traditional representation and implementation of MRFIR employ polyphase decomposition of the original filter structure, whose main purpose is to compute only the needed output at the lowest possible sampling rate. In this paper, an alternative representation and implementation technique, called TD-MRFIR (Thread Decomposition MRFIR), is presented. The basic idea is to decompose MRFIR into output computational threads, in contrast to a structural decomposition of the original filter as done in the polyphase decomposition. Each thread represents an instance of the finite convolution required to produce a single output of the MRFIR. The filter is thus viewed as a finite collection of concurrent threads. The technical details of TD-MRFIR will be explained, first showing its applicability to the implementation of downsampling, upsampling, and resampling FIR filters, and then describing a general strategy to optimally allocate the number of filter taps. A particular FPGA design of multi-stage TD-MRFIR for the L-band radar of NASA's SMAP (Soil Moisture Active Passive) instrument is demonstrated; and its implementation results in several targeted FPGA devices are summarized in terms of the functional (bit width, fixed-point error) and performance (time closure, resource usage, and power estimation) parameters.

DSP↗

Optimized FPGA Implementation of Multi-Rate FIR Filters Through Thread Decomposition

Multi-rate finite impulse response (MRFIR) filters are among the essential signal-processing components in spaceborne instruments where finite impulse response filters are often used to minimize nonlinear group delay and finite precision effects. Cascaded (multistage) designs of MRFIR filters are further used for large rate change ratio in order to lower the required throughput, while simultaneously achieving comparable or better performance than single-stage designs. Traditional representation and implementation of MRFIR employ polyphase decomposition of the original filter structure, whose main purpose is to compute only the needed output at the lowest possible sampling rate. In this innovation, an alternative representation and implementation technique called TD-MRFIR (Thread Decomposition MRFIR) is presented. The basic idea is to decompose MRFIR into output computational threads, in contrast to a structural decomposition of the original filter as done in the polyphase decomposition. A naive implementation of a decimation filter consisting of a full FIR followed by a downsampling stage is very inefficient, as most of the computations performed by the FIR state are discarded through downsampling. In fact, only 1/M of the total computations are useful (M being the decimation factor). Polyphase decomposition provides an alternative view of decimation filters, where the downsampling occurs before the FIR stage, and the outputs are viewed as the sum of M sub-filters with length of N/M taps. Although this approach leads to more efficient filter designs, in general the implementation is not straightforward if the numbers of multipliers need to be minimized. In TD-MRFIR, each thread represents an instance of the finite convolution required to produce a single output of the MRFIR. The filter is thus viewed as a finite collection of concurrent threads. Each of the threads completes when a convolution result (filter output value) is computed, and activated when the first input of the convolution becomes available. Thus, the new threads get spawned at exactly the rate of N/M, where N is the total number of taps, and M is the decimation factor. Existing threads retire at the same rate of N/M. The implementation of an MRFIR is thus transformed into a problem to statically schedule the minimum number of multipliers such that all threads can be completed on time. Solving the static scheduling problem is rather straightforward if one examines the Thread Decomposition Diagram, which is a table-like diagram that has rows representing computation threads and columns representing time. The control logic of the MRFIR can be implemented using simple counters. Instead of decomposing MRFIRs into subfilters as suggested by polyphase decomposition, the thread decomposition diagrams transform the problem into a familiar one of static scheduling, which can be easily solved as the input rate is constant.

Kobayashi, Kayla N.↗

Low synchronization Gram–Schmidt and generalized minimal residual algorithms

The Gram–Schmidt process uses orthogonal projection to construct the A = QR factorization of a matrix. When Q has linearly independent columns, the operator P = I - Q(QTQ)-1QT defines an orthogonal projection onto Q⊥. In finite precision, Q loses orthogonality as the factorization progresses. A family of approximate projections is derived with the form P = I - QTQT, with correction matrix T. When T = (QTQ)-1, and T is triangular, it is postulated that the best achievable orthogonality is $\mathcal{O}(ε)\mathcal{K}(A)$. We present new variants of modified (MGS) and classical Gram–Schmidt algorithms that require one global reduction step. An interesting form of the projector leads to a compact WY representation for MGS. In particular, the inverse compact WY MGS algorithm is equivalent to a lower triangular solve. Our main contribution is to introduce a backward normalization lag into the compact WY representation, resulting in a $\mathcal{O}(ε)\mathcal{K}[r_0, AV_m])$ stable Generalized Minimal Residual Method (GMRES) algorithm that requires only one global reduce per iteration. Finally, further improvements in performance are achieved by accelerating GMRES on GPUs.

97 MATHEMATICS AND COMPUTING↗