Search NASA⌕ Search

SEARCH · Search NASA

Results for “network inference”

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 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Architectural Implications of Neural Network Inference for High Data-Rate, Low-Latency Scientific Applications

With more scientific fields relying on neural networks (NNs) to process data incoming at extreme throughputs and latencies, it is crucial to develop NNs with all their parameters stored on-chip. In many of these applications, there is not enough time to go off-chip and retrieve weights. Even more so, off-chip memory such as DRAM does not have the bandwidth required to process these NNs as fast as the data is being produced (e.g., every 25 ns). As such, these extreme latency and bandwidth requirements have architectural implications for the hardware intended to run these NNs: 1) all NN parameters must fit on-chip, and 2) codesigning custom/reconfigurable logic is often required to meet these latency and bandwidth constraints. In our work, we show that many scientific NN applications must run fully on chip, in the extreme case requiring a custom chip to meet such stringent constraints.

Weng, Olivia↗
DOE OSTI2018

Bayesian uncertainty calculation in neural network inference of ion and electron temperature profiles at W7-X

We make use of a Bayesian description of the neural network (NN) training for the calculation of the uncertainties in the NN prediction. Having uncertainties on the NN prediction allows having a quantitative measure for trusting the NN outcome and comparing it with other methods. Within the Bayesian framework, the uncertainties can be calculated under different approximations. The NN has been trained with the purpose of inferring ion and electron temperature profile from measurements of a X-ray imaging diagnostic at W7-X. The NN has been trained in such a way that it constitutes an approximation of a full Bayesian model of the diagnostic, implemented within the Minerva framework. Finally, the network has been evaluated using measured data and the uncertainties calculated under different approximations have been compared with each other, finding that neglecting the noise on the NN input can lead to an underestimation of the error bar magnitude in the range of 10%–30%.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Ps and Qs: Quantization-Aware Pruning for Efficient Low Latency Neural Network Inference

Efficient machine learning implementations optimized for inference in hardware have wide-ranging benefits, depending on the application, from lower inference latency to higher data throughput and reduced energy consumption. Two popular techniques for reducing computation in neural networks are pruning, removing insignificant synapses, and quantization, reducing the precision of the calculations. In this work, we explore the interplay between pruning and quantization during the training of neural networks for ultra low latency applications targeting high energy physics use cases. Techniques developed for this study have potential applications across many other domains. We study various configurations of pruning during quantization-aware training, which we term quantization-aware pruning, and the effect of techniques like regularization, batch normalization, and different pruning schemes on performance, computational complexity, and information content metrics. We find that quantization-aware pruning yields more computationally efficient models than either pruning or quantization alone for our task. Further, quantization-aware pruning typically performs similar to or better in terms of computational efficiency compared to other neural architecture search techniques like Bayesian optimization. Surprisingly, while networks with different training configurations can have similar performance for the benchmark application, the information content in the network can vary significantly, affecting its generalizability.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS2000

Genetic Network Inference: From Co-Expression Clustering to Reverse Engineering

Advances in molecular biological, analytical, and computational technologies are enabling us to systematically investigate the complex molecular processes underlying biological systems. In particular, using high-throughput gene expression assays, we are able to measure the output of the gene regulatory network. We aim here to review datamining and modeling approaches for conceptualizing and unraveling the functional relationships implicit in these datasets. Clustering of co-expression profiles allows us to infer shared regulatory inputs and functional pathways. We discuss various aspects of clustering, ranging from distance measures to clustering algorithms and multiple-duster memberships. More advanced analysis aims to infer causal connections between genes directly, i.e., who is regulating whom and how. We discuss several approaches to the problem of reverse engineering of genetic networks, from discrete Boolean networks, to continuous linear and non-linear models. We conclude that the combination of predictive modeling with systematic experimental verification will be required to gain a deeper insight into living organisms, therapeutic targeting, and bioengineering.

Dhaeseleer, Patrik↗
DOE OSTI2023

Few-view computed tomography reconstruction using deep neural network inference

A system for generating 2D slices of a 3D image of a target volume is provided. The system receives a target sinogram collected during a computed tomography scan of the target volume. The system inputs the target sinogram to a convolutional neural network (CNN) to generate predicted 2D slices of the 3D image. The CNN is trained using training 2D slices of training 3D images. The system initializes 2D slices to the predicted 2D slices. The system reconstructs 2D slices of the 3D image from the target sinogram and the initialized 2D slices.

Kim, Hyojin↗
DOE OSTI2024

Distributed-Memory Sparse Deep Neural Network Inference Using Global Arrays

Partitioned Global Address Space (PGAS) models exhibit tremendous promise in developing efficient and productive distributed-memory parallel applications. They have been used extensively in scientific computations due to conveniently offering a ``shared-memory''-like model and convenient interfaces that separate communication with synchronization. Traditionally, PGAS communication models have been applied to dense/contiguously distributed data, but most modern applications depict varied levels of sparsity. Existing PGAS models require certain adaptations to support distributed sparse computations, since associated computations often require matrix arithmetic, in addition to data movement. The Global Arrays toolkit from Pacific Northwest National Laboratory (PNNL) is one of the earliest PGAS models to combine one-sided data communication and distributed matrix operations and is still used in the popular NWChem quantum chemistry suite. Recently, we have expanded the Global Arrays toolkit to support common sparse operations, like sparse matrix-dense matrix multiplies (SpMM), sparse matrix-sparse matrix multiplication (SpGEMM) and Sampled Dense-Dense Matrix Multiplication (SDDMM). As it turns out, these operations are the bedrock of sparse Deep Learning (DL); sparse deep neural networks and Graph Neural Networks (GNNs) have gained increasing attention recently in achieving speedups on training and inference with reduced memory footprints. Unlike scientific applications in High Performance Computing (HPC), modern (distributed-memory capable) DL toolkits often rely on non-standardized and closed-source vendor software optimizations, creating challenges in software-hardware co-design at scale. Our goal is to support a variety of distributed-memory sparse matrix operations and helper functions in the newly created Sparse Global Arrays (SGA), such that it is possible to build portable and productive Machine Learning scenarios for algorithm/software and hardware codesign purposes. Contemporary data-parallel schemes for training/inference are undergoing a major overhaul since model replication limits scalability and causes resource inefficiencies. As such, we have adopted tensor parallelism in decomposing the model and inputs, to mitigate memory issues. Current implementation is built on top of MPI and uses CPUs to maximize the portability across the platforms.

Distributed computing, machine learning↗
DOE OSTI2019

O3BNN: An Out-Of-Order Architecture for High-Performance Binarized Neural Network Inference with Fine-Grained Pruning

In this work, we demonstrate that the highly-condensed BNN model can be shrunk significantly further by dynamically pruning irregular redundant edges. Based on two new observations on BNN-specific properties, our out-of-order (OoO) architecture – O3BNN, can curtail remaining edge evaluation in cases where the binary output of a neuron can be determined early. Similar to Instruction-Level-Parallelism (ILP), these fine-grained, irregular, runtime pruning opportunities are traditionally presumed to be di cult to exploit. We evaluate our design on an FPGA platform using three well-known networks, including VggNet-16, AlexNet for ImageNet, and a VGG-like network for Cifar-10. Results show that our out-of-order approach can prune 27%, 16%, and 42% of the operations for the three networks respectively, without any accuracy loss, leading to at least 1.7×, 1.5×, and 2.1× speedups over state-of-the-art implementations on FPGA/GPU/CPU BNN implementations. Our approach is inference runtime pruning, so no retrain or fine-tuning is needed. We demonstrate our design on an FPGA platform. However, this is only for showcasing the method; the approach does not rely on any FPGA-specific features and can thus be adopted by other devices as well.

Geng, Tong↗
DOE OSTI2023

HOLISMOKES IX. Neural network inference of strong-lens parameters and uncertainties from ground-based images

Modeling of strong gravitational lenses is a necessity for further applications in astrophysics and cosmology. With the large number of detections in current and upcoming surveys, such as the Rubin Legacy Survey of Space and Time (LSST), it is pertinent to investigate automated and fast analysis techniques beyond the traditional and time-consuming Markov chain Monte Carlo sampling methods. Building upon our (simple) convolutional neural network (CNN), we present here another CNN, specifically a residual neural network (ResNet), that predicts the five mass parameters of a singular isothermal ellipsoid (SIE) profile (lens center x and y, ellipticity e x and e y , Einstein radius θ E ) and the external shear (γ ext, 1 , γ ext, 2 ) from ground-based imaging data. In contrast to our previous CNN, this ResNet further predicts the 1σ uncertainty for each parameter. To train our network, we use our improved pipeline to simulate lens images using real images of galaxies from the Hyper Suprime-Cam Survey (HSC) and from the Hubble Ultra Deep Field as lens galaxies and background sources, respectively. We find very good recoveries overall for the SIE parameters, especially for the lens center in comparison to our previous CNN, while significant differences remain in predicting the external shear. From our multiple tests, it appears that most likely the low ground-based image resolution is the limiting factor in predicting the external shear. Given the run time of milli-seconds per system, our network is perfectly suited to quickly predict the next appearing image and time delays of lensed transients. Therefore, we use the network-predicted mass model to estimate these quantities and compare to those values obtained from our simulations. Unfortunately, the achieved precision allows only a first-order estimate of time delays on real lens systems and requires further refinement through follow-up modeling. Nonetheless, our ResNet is able to predict the SIE and shear parameter values in fractions of a second on a single CPU, meaning that we are able to efficiently process the huge amount of galaxy-scale lenses expected in the near future.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

A Framework for Neural Network Inference on FPGA-Centric SmartNICs

FPGA-based SmartNICs offer great potential to significantly improve the performance of high-performance computing and warehouse data processing by tightly coupling support for reconfigurable data-intensive computation with cross-node communication, thereby mitigating the von Neumann bottleneck. Existing work, however, has been generally been limited in that it assumes an accelerator model where kernels are offloaded to SmartNICs, but most control tasks are left to the CPUs. This leads to frequent waiting, inferior performance, and scaling challenges. In this work, we propose a new distributive data-centric computing framework, named FCsN, for reconfigurable SmartNIC-based systems. Through a lightweight task circulation execution model and its implementation architecture, FCsN allows the complete detaching of kernel execution, control logic, system scheduling, and network communication to the SmarNICs. This boosts performance by: (i) avoiding the control dependency with CPUs and (ii) supporting streaming kernel execution and network communication at line rate and in a very fine-grained manner. We demonstrate the efficiency and flexibility of FCsN using various types of neural network applications including graph neural networks; as these last are both irregular and data intensive they offer an especially robust demonstration. Evaluations using commonly-used neural network models and graph datasets show that a system with the support of FCsN can achieve, on average, 144 speedups over the MPI-based standard CPU baselines.

Guo, Anqi↗
DOE OSTI2023

AutoReP: Automatic ReLU Replacement for Fast Private Network Inference

The proliferation of the Machine-Learning-As-A-Service (MLaaS) market has brought to light a number of clients’ data privacy and security concerns. One promising solution is private inference (PI) techniques using cryptographic primitives. These techniques often come with high computation and communication overhead associated with the non-linear operator such as ReLU. Several approaches have been developed in reducing the number of ReLU operations, however, they either require a heuristic threshold selection or introduce significant accuracy drop. This work presents AutoReP, a gradient-based framework for non-linear operators reduction that aims to mitigate these concerns from a systematic perspective. AutoReP automates the process of discrete selection of ReLU and polynomial functions on neurons to accelerate PI applications. We also introduce distribution-aware polynomial approximation (DaPa) to accurately approximate ReLUs under given distribution, preserving model expressivity. Our experimental results demonstrate significant accuracy improvements of 6.12% (94.31%, 12.9K ReLU budget, CIFAR-10), 8.39% (74.92%, 12.9K ReLU budget, CIFAR-100), and 9.45% (63.69%, 55K ReLU budget, Tiny-ImageNet) over current state-of-the-art methods, e.g., SNL. Morever, AutoReP is applied to EfficientNet-B2 on ImageNet dataset, and achieved 75.55% accuracy with 176.1 × ReLU budget reduction.

Peng, Hongwu↗
DOE OSTI2018

Fast inference of deep neural networks in FPGAs for particle physics

Here, recent results at the Large Hadron Collider (LHC) have pointed to enhanced physics capabilities through the improvement of the real-time event processing techniques. Machine learning methods are ubiquitous and have proven to be very powerful in LHC physics, and particle physics as a whole. However, exploration of the use of such techniques in low-latency, low-power FPGA (Field Programmable Gate Array) hardware has only just begun. FPGA-based trigger and data acquisition systems have extremely low, sub-microsecond latency requirements that are unique to particle physics. We present a case study for neural network inference in FPGAs focusing on a classifier for jet substructure which would enable, among many other physics scenarios, searches for new dark sector particles and novel measurements of the Higgs boson. While we focus on a specific example, the lessons are far-reaching. A companion compiler package for this work is developed based on High-Level Synthesis (HLS) called hls4ml to build machine learning models in FPGAs. The use of HLS increases accessibility across a broad user community and allows for a drastic decrease in firmware development time. We map out FPGA resource usage and latency versus neural network hyperparameters to identify the problems in particle physics that would benefit from performing neural network inference with FPGAs. For our example jet substructure model, we fit well within the available resources of modern FPGAs with a latency on the scale of 100 ns.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2019

Unified feature association networks through integration of transcriptomic and proteomic data

High-throughput multi-omics studies and corresponding network analyses of multi-omic data have rapidly expanded their impact over the last 10 years. As biological features of different types (e.g. transcripts, proteins, metabolites) interact within cellular systems, the greatest amount of knowledge can be gained from networks that incorporate multiple types of -omic data. However, biological and technical sources of variation diminish the ability to detect cross-type associations, yielding networks dominated by communities comprised of nodes of the same type. We describe here network building methods that can maximize edges between nodes of different data types leading to integrated networks. We systematically rank several network inference methods and demonstrate that, in many cases, using a random forest method, GENIE3, produces the most integrated networks. This increase in integration does not come at the cost of accuracy as GENIE3 produces networks of approximately the same quality as the other network inference methods tested here. Using GENIE3, we also infer networks representing antibody-mediated Dengue virus cell invasion and receptor-mediated Dengue virus invasion. A number of functional pathways showed centrality differences between the two networks including genes responding to both GM-CSF and IL-4, which had a higher centrality value in an antibody-mediated vs. receptor-mediated Dengue network. Because a biological system involves the interplay of many different types of molecules, incorporating multiple data types into networks will improve their use as models of biological systems. The methods explored here are some of the first to specifically highlight and address the challenges associated with how such multi-omic networks can be assembled and how the greatest number of interactions can be inferred from different data types. The resulting networks can lead to the discovery of new host response patterns and interactions during viral infection, generate new hypotheses of pathogenic mechanisms and confirm disease associated mechanisms of disease.

59 BASIC BIOLOGICAL SCIENCES↗