Search NASA⌕ Search

Engineering topics

Mahoney, Michael W.

Publications and source records attributed to Mahoney, Michael W..

End-to-end codesign of Hessian-aware quantized neural networks for FPGAs

Here, we develop an end-to-end workflow for the training and implementation of co-designed neural networks (NNs) for efficient field-programmable gate array (FPGA) hardware. Our approach leverages Hessian-aware quantization of NNs, the Quantized Open Neural Network Exchange intermediate representation, and the hls4ml tool flow for transpiling NNs into FPGA firmware. This makes efficient NN implementations in hardware accessible to nonexperts in a single open sourced workflow that can be deployed for real-time machine-learning applications in a wide range of scientific and industrial settings. We demonstrate the workflow in a particle physics application involving trigger decisions that must operate at the 40-MHz collision rate of the CERN Large Hadron Collider (LHC). Given the high collision rate, all data processing must be implemented on FPGA hardware within the strict area and latency requirements. Based on these constraints, we implement an optimized mixed-precision NN classifier for high-momentum particle jets in simulated LHC proton-proton collisions.

47 OTHER INSTRUMENTATION↗

Reliable edge machine learning hardware for scientific applications

Extreme data rate scientific experiments create massive amounts of data that require efficient ML edge processing. This leads to unique validation challenges for VLSI implementations of ML algorithms: enabling bit-accurate functional simulations for performance validation in experimental software frameworks, verifying those ML models are robust under extreme quantization and pruning, and enabling ultra-fine-grained model inspection for efficient fault tolerance. We discuss approaches to developing and validating reliable algorithms at the scientific edge under such strict latency, resource, power, and area requirements in extreme experimental environments. We study metrics for developing robust algorithms, present preliminary results and mitigation strategies, and conclude with an outlook of these and future directions of research towards the longer-term goal of developing autonomous scientific experimentation methods for accelerated scientific discovery.

Baldi, Tommaso↗

Extensions to the SENSEI In situ Framework for Heterogeneous Architectures

The proliferation of GPUs and accelerators in recent supercomputing systems, so called heterogeneous architectures, has led to increased complexity in execution environments and programming models as well as to deeper memory hierarchies on these systems. In this work, we discuss challenges that arise in in situ code coupling on these heterogeneous architectures. In particular, we present data and execution model extensions to the SENSEI in situ framework that are targeted at the effective use of systems with heterogeneous architectures. We then use these new data and execution model extensions to investigate several in situ placement and execution configurations and to analyze the impact these choices have on overall performance.

Loring, Burlen↗

Learning continuous models for continuous physics

Abstract Dynamical systems that evolve continuously over time are ubiquitous throughout science and engineering. Machine learning (ML) provides data-driven approaches to model and predict the dynamics of such systems. A core issue with this approach is that ML models are typically trained on discrete data, using ML methodologies that are not aware of underlying continuity properties. This results in models that often do not capture any underlying continuous dynamics—either of the system of interest, or indeed of any related system. To address this challenge, we develop a convergence test based on numerical analysis theory. Our test verifies whether a model has learned a function that accurately approximates an underlying continuous dynamics. Models that fail this test fail to capture relevant dynamics, rendering them of limited utility for many scientific prediction tasks; while models that pass this test enable both better interpolation and better extrapolation in multiple ways. Our results illustrate how principled numerical analysis methods can be coupled with existing ML training/testing methodologies to validate models for science and engineering applications.

97 MATHEMATICS AND COMPUTING↗

Inexact Newton-CG algorithms with complexity guarantees

Abstract We consider variants of a recently developed Newton-CG algorithm for nonconvex problems (Royer, C. W. & Wright, S. J. (2018) Complexity analysis of second-order line-search algorithms for smooth nonconvex optimization. SIAM J. Optim., 28, 1448–1477) in which inexact estimates of the gradient and the Hessian information are used for various steps. Under certain conditions on the inexactness measures, we derive iteration complexity bounds for achieving $\epsilon $-approximate second-order optimality that match best-known lower bounds. Our inexactness condition on the gradient is adaptive, allowing for crude accuracy in regions with large gradients. We describe two variants of our approach, one in which the step size along the computed search direction is chosen adaptively, and another in which the step size is pre-defined. To obtain second-order optimality, our algorithms will make use of a negative curvature direction on some steps. These directions can be obtained, with high probability, using the randomized Lanczos algorithm. In this sense, all of our results hold with high probability over the run of the algorithm. We evaluate the performance of our proposed algorithms empirically on several machine learning models. Our approach is a first attempt to introduce inexact Hessian and/or gradient information into the Newton-CG algorithm of Royer & Wright (2018, Complexity analysis of second-order line-search algorithms for smooth nonconvex optimization. SIAM J. Optim., 28, 1448–1477).

Mathematics↗

Learning from learning machines: improving the predictive power of energy-water-land nexus models with insights from complex measured and simulated data

Focal Area(s): Insights gleaned from complex data (observed and simulated) using AI; Predictive modeling through the use of AI techniques and AI-derived model components, including physics- and knowledge-informed models; Energy-water-land nexus and integrated energy systems – models of MultiSector dynamics. Science Challenge: Scientific communities are in need of tools for the computational integration of physics-based models, experimental data, and empirical/observational studies across a broad range of temporal and spatial scales to explore and meet EESSD grand challenges. We require AI algorithms for the discovery of process-drivers in the Earth-energy-human system and to link them with complex measured and simulated data. We aim to understand the energy-water-land nexus, particularly under extreme forcing scenarios and rare events, leveraging scale-aware AI process models, including probabilistic uncertainties, and benefiting from the emerging 5G-enabled landscape-scale sensing and edge computing capabilities. These models are needed to identify instabilities and tipping points that manifest extreme system behaviors with consequences for integrated energy systems and the environment. This work requires fundamental advances in uncertainty quantification, in particular, to identify and model unlikely but catastrophic outliers. Specifically, we need models that are interpretable to domain scientists and, essentially, explainable to public and private stake holders.

54 ENVIRONMENTAL SCIENCES↗

Avoiding Communication in Primal and Dual Block Coordinate Descent Methods

Primal and dual block coordinate descent methods are iterative methods for solving regularized and unregularized optimization problems. Distributed-memory parallel implementations of these methods have become popular in analyzing large machine learning datasets. However, existing implementations communicate at every iteration, which, on modern data center and supercomputing architectures, often dominates the cost of floating-point computation. Recent results on communication-avoiding Krylov subspace methods suggest that large speedups are possible by re-organizing iterative algorithms to avoid communication. We show how applying similar algorithmic transformations can lead to primal and dual block coordinate descent methods that only communicate every $s$ iterations---where $s$ is a tuning parameter---instead of every iteration for the regularized least-squares problem. We show that the communication-avoiding variants reduce the number of synchronizations by a factor of $s$ on distributed-memory parallel machines without altering the convergence rate and attain strong scaling speedups of up to $6.1\times$ over the “standard algorithm" on a Cray XC30 supercomputer.

97 MATHEMATICS AND COMPUTING↗

MAPPING THE SIMILARITIES OF SPECTRA: GLOBAL AND LOCALLY-BIASED APPROACHES TO SDSS GALAXIES

We present a novel approach to studying the diversity of galaxies. It is based on a novel spectral graph technique, that of locally-biased semi-supervised eigenvectors . Our method introduces new coordinates that summarize an entire spectrum, similar to but going well beyond the widely used Principal Component Analysis (PCA). Unlike PCA, however, this technique does not assume that the Euclidean distance between galaxy spectra is a good global measure of similarity. Instead, we relax that condition to only the most similar spectra, and we show that doing so yields more reliable results for many astronomical questions of interest. The global variant of our approach can identify very finely numerous astronomical phenomena of interest. The locally-biased variants of our basic approach enable us to explore subtle trends around a set of chosen objects. The power of the method is demonstrated in the Sloan Digital Sky Survey Main Galaxy Sample, by illustrating that the derived spectral coordinates carry an unprecedented amount of information.

79 ASTRONOMY AND ASTROPHYSICS↗

Identifying Important Ions and Positions in Mass Spectrometry Imaging Data Using CUR Matrix Decompositions

Mass spectrometry imaging enables label-free, high-resolution spatial mapping of the chemical composition of complex, biological samples. Typical experiments require selecting ions and/or positions from the images: ions for fragmentation studies to identify keystone compounds and positions for follow up validation measurements using microdissection or other orthogonal techniques. Unfortunately, with modern imaging machines, these must be selected from an overwhelming amount of raw data. Existing techniques to reduce the volume of data, the most popular of which are principle component analysis and non-negative matrix factorization, have the disadvantage that they return difficult-to-interpret linear combinations of actual data elements. In this work, we show that CX and CUR matrix decompositions can be used directly to address this selection need. CX and CUR matrix decompositions use empirical statistical leverage scores of the input data to provide provably good low-rank approximations of the measured data that are expressed in terms of actual ions and actual positions, as opposed to difficult-to-interpret eigenions and eigenpositions. We show that this leads to effective prioritization of information for both ions and positions. In particular, important ions can be found either by using the leverage scores as a ranking function and using a deterministic greedy selection algorithm or by using the leverage scores as an importance sampling distribution and using a random sampling algorithm; however, selection of important positions from the original matrix performed significantly better when they were chosen with the random sampling algorithm. Also, we show that 20 ions or 40 locations can be used to reconstruct the original matrix to a tolerance of 17% error for a widely studied image of brain lipids; and we provide a scalable implementation of this method that is applicable for analysis of the raw data where there are often more than a million rows and/or columns, which is larger than SVD-based low-rank approximation methods can handle. Lastly, these results introduce the concept of CX/CUR matrix factorizations to mass spectrometry imaging, describing their utility and illustrating principled algorithmic approaches to deal with the overwhelming amount of data generated by modern mass spectrometry imaging.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Objective identification of informative wavelength regions in galaxy spectra

Understanding the diversity in spectra is the key to determining the physical parameters of galaxies. The optical spectra of galaxies are highly convoluted with continuum and lines that are potentially sensitive to different physical parameters. Defining the wavelength regions of interest is therefore an important question. In this work, we identify informative wavelength regions in a single-burst stellar population model using the CUR Matrix Decomposition. Simulating the Lick/IDS spectrograph configuration, we recover the widely used D {sub n}(4000), Hβ, and Hδ {sub A} to be most informative. Simulating the Sloan Digital Sky Survey spectrograph configuration with a wavelength range 3450-8350 Å and a model-limited spectral resolution of 3 Å, the most informative regions are: first region—the 4000 Å break and the Hδ line; second region—the Fe-like indices; third region—the Hβ line; and fourth region—the G band and the Hγ line. A principal component analysis on the first region shows that the first eigenspectrum tells primarily the stellar age, the second eigenspectrum is related to the age-metallicity degeneracy, and the third eigenspectrum shows an anti-correlation between the strengths of the Balmer and the Ca K and H absorptions. The regions can be used to determine the stellar age and metallicity in early-type galaxies that have solar abundance ratios, no dust, and a single-burst star formation history. The region identification method can be applied to any set of spectra of the user's interest, so that we eliminate the need for a common, fixed-resolution index system. We discuss future directions in extending the current analysis to late-type galaxies. ASCII formatted tables of the regional eigenspectra are available.

79 ASTRONOMY AND ASTROPHYSICS↗