Search NASA⌕ Search

SEARCH · Search NASA

Results for “Hyperdimensional 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.

9 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Hyperdimensional computing for image classification (HDC) v1.0

This is an implementation of the hyperdimensional computing technique to classify images. It consists of a python script that trains the system for a set of images from a set of images (dataset) specified by the user. This training produces hardware configuration parameters and description vectors that are then loaded into the hardware description part of the project. The hardware description consists of hardware described in Verilog (a well known language for this purpose) that is synthesizable and can be implemented in a real chip. This hardware received the training information generated by python, and then is able to accept images to produce answers for each image on which category (class) from the pre-=trained ones the image belongs to. The hardware and python training scripts are configurable and documented. The advantage of hyperdimensional computing is its robustness to errors and the easy capability for online learning (refining the training during inference slowly over time), which this implementation supports.

Michelogiannakis, Georgios [Lawrence Berkeley Nati↗
DOE OSTI2024

Exploration of Novel Neuromorphic Methodologies for Materials Applications

Many of today's most interesting questions involve understanding and interpreting complex relationships within graph-based structures. For instance, in materials science, predicting material properties often relies on analyzing the intricate network of atomic interactions. Graph neural networks (GNNs) have emerged as a popular approach for these tasks; however, they suffer from limitations such as inefficient hardware utilization and over-smoothing. Recent advancements in neuromorphic computing offer promising solutions to these challenges. In this work, we evaluate two such neuromorphic strategies known as reservoir computing and hyperdimensional computing. We compare the performance of both approaches for bandgap classification and regression using a subset of the Materials Project dataset. Our results indicate recent advances in hyperdimensional computing can be applied effectively to better represent molecular graphs.

Gobin, Derek [George Mason University, Virginia]↗
DOE OSTI2023

Superconducting Hyperdimensional Associative Memory Circuit for Scalable Machine Learning

Here we propose a generalized architecture for the first rapid-single-flux-quantum (RSFQ) associative memory circuit. The circuit employs hyperdimensional computing (HDC), a machine learning (ML) paradigm utilizing vectors with dimensionality in the thousands to represent information. HDC designs have small memory footprints, simple computations, and simple training algorithms compared to superconducting neural network accelerators (SNNAs), making them a better option for scalable SFQ machine learning (ML) solutions. The proposed superconducting HDC (SHDC) circuit uses entirely on-chip RSFQ memory which is tightly integrated with logic, operates at 33.3 GHz, is applicable to general ML tasks, and is manufacturable at practically useful scales given current SFQ fabrication limits. Tailored to a language recognition task, SHDC consists of ~ 2-20 M Josephson junctions (JJs) and consumes up to three times less power than an analogous CMOS HDC circuit while achieving 78-84% higher throughput. SHDC is capable of outperforming the state of the art RSFQ SNNA, SuperNPU, by 48-99% for all benchmark NN architectures tested while occupying up to 90% less area and consuming up to nine times less power. To the best of the authors' knowledge, SHDC is currently the only superconducting ML approach feasible at practically useful scales for real-world ML tasks and capable of online learning.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

HDBind: encoding of molecular structure with hyperdimensional binary representations

Traditional methods for identifying “hit” molecules from a large collection of potential drug-like candidates rely on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have a significant limitation in that they require exceptional computing capabilities for even relatively small collections of molecules. Increasingly large and complex state-of-the-art deep learning approaches have gained popularity with the promise to improve the productivity of drug design, notorious for its numerous failures. However, as deep learning models increase in their size and complexity, their acceleration at the hardware level becomes more challenging. Hyperdimensional Computing (HDC) has recently gained attention in the computer hardware community due to its algorithmic simplicity relative to deep learning approaches. The HDC learning paradigm, which represents data with high-dimension binary vectors, allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas (computer vision, bioinformatics, mass spectrometery, remote sensing, edge devices, etc.). To the best of our knowledge, our work is the first to consider HDC for the task of fast and efficient screening of modern drug-like compound libraries. We also propose the first HDC graph-based encoding methods for molecular data, demonstrating consistent and substantial improvement over previous work. We compare our approaches to alternative approaches on the well-studied MoleculeNet dataset and the recently proposed LIT-PCBA dataset derived from high quality PubChem assays. We demonstrate our methods on multiple target hardware platforms, including Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs), showing at least an order of magnitude improvement in energy efficiency versus even our smallest neural network baseline model with a single hidden layer. Our work thus motivates further investigation into molecular representation learning to develop ultra-efficient pre-screening tools. We make our code publicly available at https://github.com/LLNL/hdbind.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

HD-Bind: Encoding of Molecular Structure with Low Precision, Hyperdimensional Binary Representations

Publicly available collections of drug-like molecules have grown to comprise tens of billions of compounds due to advances in combinatorial chemistry. Traditional methods for identifying "hit" molecules from a large collection of potential drug-like candidates have relied on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have the major drawback that they require exceptional computing capabilities for even relatively small collections of molecules. Hyperdimensional Computing (HDC) is a recently-proposed learning paradigm that represents data with high-dimension binary vectors; this allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas. We consider existing HDC approaches for molecular property classification and introduce two novel encodings of a commonly-used molecular representation, the extended connectivity fingerprint (ECFP). We show that HDC-based inference methods are as much as 91 times more efficient than traditional machine learning methods, and achieve an acceleration of nearly nine orders of magnitude compared to molecular docking. Our results show that HDC accelerated methods retain competitive accuracy on a number of well-studied tasks such as molecular property predictions using the MoleculeNet dataset, and bind/no-bind activity classification using the DUD-E and LIT-PCBA datasets. Our work thus motivates further investigation into molecular representation learning to develop ultraefficient pre-screening tools.

Jones, William↗
DOE OSTI2023

Modernist materials synthesis: Finding thermodynamic shortcuts with hyperdimensional chemistry

Synthesis remains a challenge for advancing materials science. A key focus of this challenge is how to enable selective synthesis, particularly as it pertains to metastable materials. This perspective addresses the question: how can “spectator” elements, such as those found in double ion exchange (metathesis) reactions, enable selective materials synthesis? By observing reaction pathways as they happen (in situ) and calculating their energetics using modern computational thermodynamics, we observe transient, crystalline intermediates that suggest that many reactions attain a local thermodynamic equilibrium dictated by local chemical potentials far before achieving a global equilibrium set by the average composition. Furthermore, using this knowledge, one can thermodynamically “shortcut” unfavorable intermediates by including additional elements beyond those of the desired target, providing access to a greater number of intermediates with advantageous energetics and selective phase nucleation. Ultimately, data-driven modeling that unites first-principles approaches with experimental insights will refine the accuracy of emerging predictive retrosynthetic models for complex materials synthesis.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Dimensionality reduction using elastic measures

With the recent surge in big data analytics for hyperdimensional data, there is a renewed interest in dimensionality reduction techniques. In order for these methods to improve performance gains and understanding of the underlying data, a proper metric needs to be identified. This step is often overlooked, and metrics are typically chosen without consideration of the underlying geometry of the data. Here, in this paper, we present a method for incorporating elastic metrics into the t-distributed stochastic neighbour embedding (t-SNE) and Uniform Manifold Approximation and Projection (UMAP). We apply our method to functional data, which is uniquely characterized by rotations, parameterization and scale. If these properties are ignored, they can lead to incorrect analysis and poor classification performance. Through our method, we demonstrate improved performance on shape identification tasks for three benchmark data sets (MPEG-7, Car data set and Plane data set of Thankoor), where we achieve 0.77, 0.95 and 1.00 F1 score, respectively.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2010

Reconciling Scientific Aspirations and Engineering Constraints for a Lunar Mission via Hyperdimensional Interpolation

Virtually every NASA space-exploration mission represents a compromise between the interests of two expert, dedicated, but very different communities: scientists, who want to go quickly to the places that interest them most and spend as much time there as possible conducting sophisticated experiments, and the engineers and designers charged with maximizing the probability that a given mission will be successful and cost-effective. Recent work at NASA's Jet Propulsion Laboratory (JPL) seeks to enhance communication between these two groups, and to help them reconcile their interests, by developing advanced modeling capabilities with which they can analyze the achievement of science goals and objectives against engineering design and operational constraints. The analyses conducted prior to this study have been point-design driven. Each analysis has been of one hypothetical case which addresses the question: Given a set of constraints, how much science can be done? But the constraints imposed by the architecture team-e.g., rover speed, time allowed for extravehicular activity (EVA), number of sites at which science experiments are to be conducted- are all in early development and carry a great deal of uncertainty. Variations can be incorporated into the analysis, and indeed that has been done in sensitivity studies designed to see which constraint variations have the greatest impact on results. But if a very large number of variations can be analyzed all at once, producing a table that includes virtually the entire trade space under consideration, then we have a tool that enables scientists and mission architects to ask the inverse question: For a given desired level of science (or any other objective), what is the range of constraints that would be needed? With this tool, mission architects could determine, for example, what combinations of rover speed, EVA duration, and other constraints produce the desired results. Further, this tool would help them identify which technology-improvement investments would be likely to produce the largest or most important return. However, the number of variations that need to be considered for such analysis quickly balloons to an unwieldy size. If three variations are considered for each of six constraints-a very modest example-there are a total of 243 variations to consider. If it takes 40 minutes to compute each variation, as it does with HURON, our automated optimization system, then it would take 162 hours or nearly 7 days of round-the-clock computing to calculate the results. Adding further constraints or variations exponentially increases the amount of time that is needed.

Weisbin, Charles R.↗