SoftMemoryBox II: A Scalable, Shared Memory Buffer Framework for Accelerating Distributed Training of Large-Scale Deep Neural Networks

Parallel training accelerates the Deep Neural Networks (DNN) training by parallel GPUs. While the in-memory data transmission becomes the cross-node network transmission due to distribution of GPUs on different nodes, which drags the training time. Most researches address it by reducing the data size on network links. However, the factor of network distance is ignored. In this paper, we construct a distributed DNN training architecture based on MapReduce. The customized scheduler is designed to make the computations nodes that finish the training closer to the nodes that storage data. At the same time, the parallel training models are synchronized by adjusting the data transmission time. The experimental results show that the shortened network distance benefits the reduced network traffic usage. The resulting data transmission time decreases the training time by at least 50% and guarantees the synchronization for the parallel training.

Deep Learning (DL) has become a prominent machine learning technique due to the availability of efficient computational resources in the form of Graphics Processing Units (GPUs), large-scale datasets and a variety of models. The newer generation of GPUs are being designed with special emphasis on optimizing performance for DL applications. Also, the availability of easy-to-use DL frameworks—like PyTorch and TensorFlow— has enhanced productivity of domain experts to work on their custom DL applications from diverse domains. However, existing Deep Neural Network (DNN) training approaches may not fully utilize the newly emerging powerful GPUs like the NVIDIA A100—this is the primary issue that we address in this paper. Our motivating analyses show that the GPU utilization on NVIDIA A100 can be as low as 43% using traditional DNN training approaches for small-to-medium DL models and input data size. This paper proposes AccDP—a data-parallel distributed DNN training approach—to accelerate GPU-based DL applications. AccDP exploits the Message Passing Interface (MPI) communication library coupled with the NVIDIA’s Multi-Process Service (MPS) to increase the amount of work assigned to parallel GPUs resulting in higher utilization of compute resources. We evaluate our proposed design on different small-to-medium DL models and input sizes on the state-of-the-art HPC clusters. By injecting more parallelism into DNN training using our approach, the evaluation shows up to 58% improvement in training performance on a single GPU and up to 62% on 16 GPUs compared to regular DNN training. Furthermore, we conduct an in-depth characterization to determine the impact of several DNN training factors and best practices—including the batch size and the number of data loading workers— to optimally utilize GPU devices. To the best of our knowledge, this is the first work that explores the use of MPS and MPI to maximize the utilization of GPUs in distributed DNN training.

Distant speech recognition (DSR) remains to be an open challenge, even for the state-of-the-art deep neural network (DNN) models. Previous work has attempted to improve DNNs under constantly distant speech. However, in real applications, the speaker-microphone distance (SMD) can be quite dynamic, varying even within a single utterance. This paper investigates how to alleviate the impact of dynamic SMD on DNN models. Our solution is to incorporate the frame-level SMD information into DNN training. Generation of the SMD information relies on a universal extractor that is learned on a meeting corpus. We study the utility of different architectures in instantiating the SMD extractor. On our target acoustic modeling task, two approaches are proposed to build distance-aware DNN models using the SMD information: simple concatenation and distance adaptive training (DAT). Our experiments show that in the simplest case, incorporating the SMD descriptors improves word error rates of DNNs by 5.6% relative. Further optimizing SMD extraction and integration results in more gains.

Machine learning has surged in popularity in recent years, with deep neural networks (DNNs) becoming a cornerstone of applications such as autonomous driving, augmented reality, and natural language processing, owing to their remarkable accuracy and scalability. As the demand for greater portability and flexibility grows, there is a strong push to optimize these applications for resource-constrained edge devices like smartphones and IoT devices.However, deploying DNN models on such devices presents significant challenges, primarily due to the large size of these models, which is a key contributor to their high performance. For instance, the DeepSpeech 2 model outperforms its predecessor, DeepSpeech 1, but its size is eight times larger, encompassing approximately 68 million parameters. Similarly, other networks designed for various tasks, including convolutional neural networks like VGG and transformer-based networks like BERT and GPT, also exhibit substantial sizes. For example, the GPT-3 model, a predecessor to ChatGPT, consists of an astonishing 175 billion parameters. Consequently, fitting DNN models onto resource-limited edge devices is a formidable challenge. My thesis addresses these challenges with a three-part approach: * High-Performance DNN Accelerators: I will develop DNN accelerators that excel in high performance, low latency, and energy efficiency through an algorithm and hardware co-design across diverse FPGA architectures (Chapter 2). * Real-Time DNN Execution: I aim to achieve exceptional accuracy and real-time execution for various large-scale DNNs on off-the-shelf mobile devices, IoT devices, and other platforms (Chapter 3). * CompVQC Framework: I will explore a systematic framework, CompVQC, to reduce the circuit length of quantum neural networks (QNNs) through model compression, which can achieve up to a 20% improvement in accuracy. This reveals that compilation-aware compression can enhance the robustness of QNNs on near-term noisy quantum devices (Chapter 4). The field of AI has recently undergone a significant transformation, ushering in the era of generative AI, where models like ChatGPT and Stable Diffusion create realistic and creative content for a wide range of applications. As generative AI becomes more prevalent, it is crucial to adapt algorithm-hardware co-design strategies to meet the specific needs of these models, particularly on edge devices. In the final chapter (Chapter 5), I will introduce my ongoing and future work focused on activating generative AI on the edge.--Author's abstract

Deep learning is being incorporated in many modern software systems. Deep learning approaches train a deep neural network (DNN) model using training examples, and then use the DNN model for prediction. While the structure of a DNN model as layers is observable, the model is treated in its entirety as a monolithic component. To change the logic implemented by the model, e.g. to add/remove logic that recognizes inputs belonging to a certain class, or to replace the logic with an alternative, the training examples need to be changed and the DNN needs to be retrained using the new set of examples. We argue that decomposing a DNN into DNN modules— akin to decomposing a monolithic software code into modules—can bring the benefits of modularity to deep learning. In this work, we develop a methodology for decomposing DNNs for multi-class problems into DNN modules. For four canonical problems, namely MNIST, EMNIST, FMNIST, and KMNIST, we demonstrate that such decomposition enables reuse of DNN modules to create different DNNs, enables replacement of one DNN module in a DNN with another without needing to retrain. The DNN models formed by composing DNN modules are at least as good as traditional monolithic DNNs in terms of test accuracy for our problems.

Automated image classification is an essential task of the computer vision field. The tagging of images into a set of predefined groups is referred to as image classification. The implementation of computer vision to automate image classification would be beneficial because manual image evaluation and identification can be time-consuming, particularly when there are many images of different classes. Deep learning approaches are proven to overperform existing machine learning techniques in many fields in recent years, and computer vision is one of the most notable examples. The very deep neural network (VDCNN) is a powerful deep learning model for image classification, and this paper examines it briefly using MNIST handwritten digit dataset. This dataset is used to prove the efficacy of very deep neural networks over other deep learning models. The proposed study aims to comprehend the very deep neural network architecture used to accomplish a handwritten digit recognition task. The feasibility of the proposed model is evaluated using mean accuracy, validation accuracy, and standard deviation. The study results of the very deep neural network model are compared to a convolutional neural network and convolutional neural network with batch normalization. According to the results of the comparison study, very deep neural networks achieve high accuracy of 99.1% for a handwritten dataset. The outcome of the proposed work is used to interpret how well a very deep neural network performs when compared to the other two models of deep neural networks. This proposed architecture may be used to automate the classification of handwritten digits dataset.

An imbalanced dataset is a significant challenge when training a deep neural network (DNN) model for deep learning problems, such as weeds classification. An imbalanced dataset may result in a model that behaves robustly on major classes and is overly sensitive to minor classes. This article proposes a yielding multi-fold training (YMufT) strategy to train a DNN model on an imbalanced dataset. This strategy reduces the bias in training through a min-class-max-bound procedure (MCMB), which divides samples in the training set into multiple folds. The model is consecutively trained on each one of these folds. In practice, we experiment with our proposed strategy on two small (PlantSeedlings, small PlantVillage) and two large (Chonnam National University (CNU), large PlantVillage) weeds datasets. With the same training configurations and approximate training steps used in conventional training methods, YMufT helps the DNN model to converge faster, thus requiring less training time. Despite a slight decrease in accuracy on the large dataset, YMufT increases the F1 score in the NASNet model to 0.9708 on the CNU dataset and 0.9928 when using the Mobilenet model training on the large PlantVillage dataset. YMufT shows outstanding performance in both accuracy and F1 score on small datasets, with values of (0.9981, 0.9970) using the Mobilenet model for training on small PlantVillage dataset and (0.9718, 0.9689) using Resnet to train on the PlantSeedlings dataset. Grad-CAM visualization shows that conventional training methods mainly concentrate on high-level features and may capture insignificant features. In contrast, YMufT guides the model to capture essential features on the leaf surface and properly localize the weeds targets.

COVID-19 is most commonly diagnosed using a testing kit but chest X-rays and computed tomography (CT) scan images have a potential role in COVID-19 diagnosis. Currently, CT diagnosis systems based on Artificial intelligence (AI) models have been used in some countries. Previous research studies used complex neural networks, which led to difficulty in network training and high computation rates. Hence, in this study, we developed the 6-layer Deep Neural Network (DNN) model for COVID-19 diagnosis based on CT scan images. The proposed DNN model is generated to improve accurate diagnostics for classifying sick and healthy persons. Also, other classification models, such as decision trees, random forests and standard neural networks, have been investigated. One of the main contributions of this study is the use of the global feature extractor operator for feature extraction from the images. Furthermore, the 10-fold cross-validation technique is utilized for partitioning the data into training, testing and validation. During the DNN training, the model is generated without dropping out of neurons in the layers. The experimental results of the lightweight DNN model demonstrated that this model has the best accuracy of 96.71% compared to the previous classification models for COVID-19 diagnosis.

Deep Neural Network (DNN) models have achieved excellent performance in many inference tasks and have been widely used in many intelligent applications. However, DNN models often require a lot of computational resources to complete the inference tasks, which hinders the deployment of such models to resource‐constrained edge devices. In order to extend the application scenarios of DNN models, the edge‐cloud collaborative inference methods, represented by model partition, have attracted much research attention in recent years. In scenarios that have multiple edge devices deployed, the edge‐cloud collaborative inference method requires partial migration of tasks, but traditional scheduling methods only migrate tasks at the task level. In this paper, we propose two task scheduling methods, which can solve the problem of partial migration of tasks in multiedge scenarios. The first scheduling method is based on the optimal cutting of a single DNN. The cutting positions of all the models are the same, regardless of the influence of external factors. This method is suitable for chain and directed acyclic graph‐ (DAG‐) type DNNs. The second scheduling method takes external factors such as congestion and queuing delay at the cloud side into consideration, which dynamically selects the cutting position of each DNN to optimize the overall delay and thus is applicable to chain DNN models. The experimental results show that, compared with the baseline method, our proposed scheduling method can reduce the delay by up to 6.48x.

Recently deep-learning (DL) techniques have been widely adopted in side-channel power analysis. A DL-assisted SCA generally consists of two phases: a deep neural network (DNN) training phase and a follow-on attack phase using the trained DNN. However, currently the two phases are not well aligned, as there is no conclusion on what metric used in the training can result in the most effective attack in the second phase. When traditional loss functions such as negative log-likelihood (NLL) are used in training a DNN, the trained model does not yield optimal follow-on attack. Recently some information theoretical SCA leakage metrics are proposed, either as the validation metric to stop the DNN training with traditional loss functions, or as both the validation metric and the training loss function. None of those proposed metrics, however, directly measures the SCA effectiveness. We propose to conduct DNN training directly with a common SCA effectiveness metric, Guessing Entropy (GE). We overcome the prior practical difficulty of using GE in DNN training by utilizing the GEEA estimation algorithm introduced in CHES 2020. We show that using GEEA as either the validation metric or the loss function produces DNN models that lead to much more effective follow-on attacks. Our work consolidates the DL-assisted SCA framework with a consistent metric, which shows great potential to be adopted as the universal SCA-oriented DNN training framework.

Deep neural network (DNN) models typically have many hyperparameters that can be configured to achieve optimal performance on a particular dataset. Practitioners usually tune the hyperparameters of their DNN models by training a number of trial models with different configurations of the hyperparameters, to find the optimal hyperparameter configuration that maximizes the training accuracy or minimizes the training loss. As such hyperparameter tuning usually focuses on the model accuracy or the loss function, it is not clear and remains under-explored how the process impacts other performance properties of DNN models, such as inference latency and model size. On the other hand, standard DNN models are often large in size and computing-intensive, prohibiting them from being directly deployed in resource-bounded environments such as mobile devices and Internet of Things (IoT) devices. To tackle this problem, various model optimization techniques (e.g., pruning or quantization) are proposed to make DNN models smaller and less computing-intensive so that they are better suited for resource-bounded environments. However, it is neither clear how the model optimization techniques impact other performance properties of DNN models such as inference latency and battery consumption, nor how the model optimization techniques impact the effect of hyperparameter tuning (i.e., the compounding effect). Therefore, in this paper, we perform a comprehensive study on four representative and widely-adopted DNN models, i.e., CNN image classification , Resnet-50 , CNN text classification , and LSTM sentiment classification , to investigate how different DNN model hyperparameters affect the standard DNN models, as well as how the hyperparameter tuning combined with model optimization affect the optimized DNN models, in terms of various performance properties (e.g., inference latency or battery consumption). Our empirical results indicate that tuning specific hyperparameters has heterogeneous impact on the performance of DNN models across different models and different performance properties. In particular, although the top tuned DNN models usually have very similar accuracy, they may have significantly different performance in terms of other aspects (e.g., inference latency). We also observe that model optimization has a confounding effect on the impact of hyperparameters on DNN model performance. For example, two sets of hyperparameters may result in standard models with similar performance but their performance may become significantly different after they are optimized and deployed on the mobile device. Our findings highlight that practitioners can benefit from paying attention to a variety of performance properties and the confounding effect of model optimization when tuning and optimizing their DNN models.

Deep neural network modeling of unknown partial differential equations in nodal space

As training a high-performance deep neural network (DNN) model requires a large amount of data, powerful computing resources and expert knowledge, protecting well-trained DNN models from intellectual property (IP) infringement has raised serious concerns in recent years. Most existing methods using DNN watermarks to verify the ownership of the models after IP infringement occurs, which is reactive in the sense that they cannot prevent unauthorized users from using the model in the first place. Different from these methods, in this article, we propose an active authorization control and user’s fingerprint tracking method for the IP protection of DNN models by utilizing sample-specific backdoor attack. The proposed method inversely and multiplely exploits sample-specific trigger as the key to implement authorization control for DNN model, in which the generated triggers are imperceptible and sample-specific for clean images. Specifically, a U-Net model is used to generate backdoor instances. Then, the target model is trained on the clean images and backdoor instances, which are inversely labeled as wrong classes and correct classes, respectively. Only authorized users can use the target model normally by pre-processing the clean images through the U-Net model. Moreover, the images processed by the U-Net model will contain unique fingerprint that can be extracted to verify and track the corresponding user’s identity. This article is the first work that utilizes the sample-specific backdoor attack to implement active authorization control and user’s fingerprint management for DNN model under black-box scenarios. Extensive experimental results on ImageNet dataset and YouTube Aligned Face dataset demonstrate that the proposed method is effective in protecting the DNN model from unauthorized usage. Specifically, the protected model has a low inference accuracy (1.00%) for unauthorized users, while maintaining a normal inference accuracy (97.67%) for authorized users. Besides, the proposed method can achieve 100% fingerprint tracking success rates on both the ImageNet and YouTube Aligned Face datasets. Moreover, it is demonstrated that the proposed method is robust against fine-tuning attack, pruning attack, pruning attack with retraining, reverse-engineering attack, adaptive attack, and JPEG compression attack. The code is available at https://github.com/nuaaaisec/SSAT .

Accurate physical layer modeling for optical fiber communication systems is one of the key aspects of implementing an intelligent optical network. In this paper, we propose and study a data-driven model enabled by digital signal processing (DSP) methods for predicting the electrical and optical behavior of an intensity modulation direct detection (IM/DD) system based on an electro-absorption modulated laser (EML). This model captures both the linear and nonlinear characteristics of an IM/DD system with an adaptive finite impulse response (FIR) filter and a look-up table (LUT). Similar to conventional supervised machine learning problems, the proposed model is measurement-informed by iterating through a PAM-M signal training dataset. Additionally, we demonstrate that the proposed model parameters can be generalized by applying 2D piecewise bilinear interpolation across different driving voltages and received optical powers. To highlight the model advantages, we benchmark the proposed model against an existing equivalent circuit model and a deep neural network (DNN) model using multiple performance metrics, including signal root mean square error (RMSE), bit error rate (BER), training time, and model complexity. We demonstrate that the proposed FIR-LUT model outperforms the conventional equivalent circuit model using the PAM-2 dataset by achieving 0.021 sample RMSE and 0.018 symbol RMSE. Similarly, a 0.020 sample RMSE and 0.015 symbol RMSE are achieved with the PAM-4 dataset. Compared with the DNN alternative, the proposed FIR-LUT model delivers comparable accuracy but significantly reduced model complexity. The FIR-LUT model achieves <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$34\times$</tex-math></inline-formula> PAM-2 and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$24\times$</tex-math></inline-formula> PAM-4 training phase speed-up and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$486\times$</tex-math></inline-formula> PAM-2 and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$283\times$</tex-math></inline-formula> PAM-4 inference phase speed-up when compared with the DNN model.

The scale of model parameters and datasets is rapidly growing for high accuracy in various areas. To train a large-scale deep neural network (DNN) model, a huge amount of computation and memory is required; therefore, a parallelization technique for training large-scale DNN models has attracted attention. A number of approaches have been proposed to parallelize large-scale DNN models, but these schemes lack scalability because of their long communication time and limited worker memory. They often sacrifice accuracy to reduce communication time. In this work, we proposed an efficient parallelism strategy named group hybrid parallelism (GHP) to minimize the training time without any accuracy loss. Two key ideas inspired our approach. First, grouping workers and training them by groups reduces unnecessary communication overhead among workers. It saves a huge amount of network resources in the course of training large-scale networks. Second, mixing data and model parallelism can reduce communication time and mitigate the worker memory issue. Data and model paralleism are complementary to each other so the training time can be enhanced when they are combined. We analyzed the training time model of the data and model parallelism, and based on the training time model, we demonstrated the heuristics that determine the parallelization strategy for minimizing training time. We evaluated group hybrid parallelism in comparison with existing parallelism schemes, and our experimental results show that group hybrid parallelism outperforms them.