High-performance Graphics Processing Units Research Articles

CuGBasis: High-performance CUDA/Python library for efficient computation of quantum chemistry density-based descriptors for larger systems.

CuGBasis is a free and open-source CUDA®/Python library for efficient computation of scalar, vector, and matrix quantities crucial for the post-processing of electronic structure calculations. CuGBasis integrates high-performance Graphical Processing Unit (GPU) computing with the ease and flexibility of Python programming, making it compatible with a vast ecosystem of libraries. We showcase its utility as a Python library and demonstrate its seamless interoperability with existing Python software to gain chemical insight from quantum chemistry calculations. Leveraging GPU-accelerated code, cuGBasis exhibits remarkable performance, making it highly applicable to larger systems or large databases. Our benchmarks reveal a 100-fold performance gain compared to alternative software packages, including serial/multi-threaded Central Processing Unit and GPU implementations. This paper outlines various features and computational strategies that lead to cuGBasis's enhanced performance, guiding developers of GPU-accelerated code.

Efficient and lightweight convolutional neural network architecture search methods for object classification

Determining the architecture of deep learning models is a complex task. Several automated search techniques have been proposed, but these methods typically require high-performance graphics processing units (GPUs), manual parameter adjustments, and specific training approaches. This study introduces an efficient, lightweight convolutional neural network architecture search approach tailored for object classification. It features an optimized search space design and a novel controller design.This study introduces a refined search space design incorporating optimizations in both spatial and operational aspects. The focus is on the synergistic integration of convolutional units, dimension reduction units, and the stacking of Convolutional Neural Network (CNN) architectures. To enhance the search space, ShuffleNet modules are integrated, reducing the number of parameters and training time. Additionally, BlurPool is implemented in the dimension reduction unit operation to achieve translational invariance, alleviate the gradient vanishing problem, and optimize unit compositions. Moreover, an innovative controller model, Stage LSTM, is proposed based on Long Short-Term Memory (LSTM) to generate lightweight architectural sequences. In conclusion, the refined search space design and the Stage LSTM controller model are synergistically combined to establish an efficient and lightweight architecture search technique termed Stage and Lightweight Network Architecture Search (SLNAS).The experimental results highlight the superior performance of the optimized search space design, primarily when implemented with the Stage LSTM controller model. This approach shows significantly improved accuracy and stability compared to random, traditional LSTM, and Genetic Algorithm (GA) controller models, with statistically significant differences. Notably, the Stage LSTM controller excels in accuracy while producing models with fewer parameters within the expanded architecture search space.The study adopts the Stage LSTM controller model due to its ability to approximate optimal sequence structures, particularly when combined with the optimized search space design, referred to as SLNAS. SLNAS's performance is evaluated through experiments and comparisons with other Neural Architecture Search (NAS) and object classification methods from different researchers. These experiments consider model parameters, hardware resources, model stability, and multiple datasets. The results show that SLNAS achieves a low error rate of 2.86 % on the CIFAR-10 dataset after just 0.2 days of architecture search, matching the performance of manually designed models but using only 2 % of the parameters. SLNAS consistently demonstrates robust performance across various image classification domains, with an approximate parameter count 700,000.To summarize, SLNAS emerges as a highly effective automated network architecture search method tailored for image classification. It streamlines the model design process, making it accessible to researchers without specialized knowledge in deep learning. Optimizing this method unlocks the full potential of deep learning across diverse research areas. Interested parties can publicly access the source code and pre-trained models through the following link: https://github.com/huanyu-chen/LNASG-and-SLNAS-model.

Hand-held GPU accelerated device for multiclass classification of X-ray images using CNN model

Chest X-ray (CXR) images are the primary investigation aid for many lung diseases and their follow-ups. For diagnosis of SARS-CoV-2, RT–PCR test and chest Computed Tomography (CT) are commonly used but both face false negatives for ruling out the infection. So, there is a demanding need for developing a system combined with Artificial Intelligence (AI) and CXR imaging to detect COVID-19 patients to avoid its spread. Here, a robust and efficient handheld device is proposed. It uses the computational power of the Graphics Processing Unit (GPU) and pre-trained deep learning models for analyzing the CXR images. A Resnet-50 CNN model is deployed on an NVIDIA Jetson Nano GPU module for the real-time classification of COVID-19, Tuberculosis, and Normal using CXR images. The device can perform real-time classification of CXR images from a portable X-ray machine and classify the image into one of the above categories. For the extensive training, a database of 680 COVID-19, 1230 Tuberculosis, and 1050 normal CXR images are extracted by combining several global databases like Kaggle, SIRM, RSNA, and Radiopaedia. The classification accuracy, precision, and loss rate were 0.9879, 0.9758, and 0.0196 respectively and our model would improve with larger data sets. The highly accurate and high-performance GPU device significantly plays a far-reaching role in COVID-19 diagnosis using Chest X-ray, which could be beneficial to triage the health system and to combat the outbreak of COVID-19.

High performance GPU graphics API abstraction layer in C# for real-time graphics

Real-Time rendering is the technique which allows us to have graphical applications in our everyday life, whether it is a 3D game or a tool with graphical user interface. Nowadays graphics rendering is handled by the GPU (Graphics Processing Unit) in our device. There are many layers of abstraction above the programming of GPUs through libraries and graphics engines, though the most low-level way of accessing a GPU in user-mode applications is using a Graphics API. Due to the need of high performance and low-level capabilities usually these APIs are used from C or C++, but we realized the need to utilize these APIs in higher level languages as well.In our approach we're using the .NET C# language for developing multi-platform real-time graphical applications instead of the C or C++ languages. Using the modern .NET environment, we're able to use Graphics APIs for rendering onto common .NET UI Frameworks while consuming all our previously implemented C# libraries and .NET technologies in the same application. To maintain compatibility with multiple platforms we're developing a library system allowing the use different Graphics APIs from the same C# source-code. The library system contains a Graphics API abstraction layer with multiple Graphics API implementations of this layer in C# and a C# to shader language compiler for cross-API shader development in C#.In this paper, we're proposing our considerations for implementing a library to be able to use the Vulkan and OpenGL APIs through a single C# codebase. We provide solutions for multi-platform rendering and dealing with the low-level challenges of using the two deeply different APIs, while maintaining performance capable to do real-time rendering.

Accelerating Static Timing Analysis Using CPU–GPU Heterogeneous Parallelism

Static timing analysis (STA) is an essential yet time-consuming task during the circuit design flow to ensure the correctness and performance of the design. Thanks to the advancement of general-purpose computing on graphics processing units (GPUs), new possibilities and challenges have arisen for boosting the performance of STA. In this work, we present an efficient and holistic GPU-accelerated STA engine. We accelerate major STA tasks, including levelization, delay computation, graph propagation, and multi-corner analysis, by developing high-performance GPU kernels and data structures. By dividing the STA workloads into CPU-GPU concurrent tasks with managed dependencies, our acceleration framework supports versatile incremental updates. Furthermore, we have extended our approach to multi-corner analysis by exploring a large amount of corner-level data parallelism using GPU computing. Our implementation based on the open-source STA engine OpenTimer has achieved up to 4.07× speed-up on single corner analysis, and up to 25.67× speed-up on multi-corner analysis on TAU 2015 contest designs and a 14nm technology.

DEM study of the stress fields around the closed-ended displacement pile driven in sand

Three-dimensional discrete element method (DEM) is employed to model the published calibration chamber test of a closed-ended displacement pile driven in sand. To achieve large chamber-to-pile and pile-to-particle size ratios for unbiased predictions, a high-performance graphics processing unit-powered DEM code is utilised in conjunction with the axisymmetric assumption and particle refinement method. The DEM parameters are first calibrated against the drained triaxial compression test data on Fontainebleau sand. Then, the DEM modelling of pile driving demonstrates quantitative agreement with the experimental counterpart in terms of cone resistance, stress contours, and distributions.

Democratization of deep learning for segmenting cartilage from MRIs of human knees: Application to data from the osteoarthritis initiative.

In this study, we aimed to democratize access to convolutional neural networks (CNN) for segmenting cartilage volumes, generating state-of-the-art results for specialized, real-world applications in hospitals and research. Segmentation of cross-sectional and/or longitudinal magnetic resonance (MR) imagesof articular cartilage facilitates both clinical management of joint damage/disease and fundamental research. Manual delineation of such images is a time-consuming task susceptible to high intra- and interoperator variability and prone to errors. Thus, enabling reliable and efficient analyses of MRIs of cartilage requires automated segmentation of cartilage volumes. Two main limitations arise in the development of hospital- or population-specific deep learning (DL) models for image segmentation: specialized knowledge and specialized hardware. We present a relatively easy and accessible implementation of a DL model to automatically segment MRIs of human knees with state-of-the-art accuracy. In representative examples, we trained CNN models in 6-8 hand obtained results quantitatively comparable to state-of-the-art for every anatomical structure. We established and evaluated our methods using two publicly available MRI data sets originating from the Osteoarthritis Initiative, Stryker Imorphics, and Zuse Institute Berlin (ZIB), as representative test cases. We use Google Colabfor editing and adapting the Python codes and selecting the runtime environment leveraging high-performance graphical processing units. We designed our solution for novice users to apply to any data set with relatively few adaptations requiring only basic programming skills. To facilitate the adoption of our methods, we provide a complete guideline for using our methods and software, as well as the software tools themselves.Clinical significance:We establish and detail methods that clinical personal can apply to create their own DLmodels without specialized knowledge of DL nor specialized hardware/infrastructure and obtain results comparable with the state-of-the-art to facilitate both clinical management of joint damage/disease and fundamental research.

Computing efficiency of XBeach hydro- and wave dynamics on Graphics Processing Units (GPUs)

Numerical prediction of coastal inundation can be complex due to the multiple physical processes involved and typically requires two-dimensional numerical model extents, particularly in areas with complex along-shore morphology. Such model domains often incur relatively high computational expense. Recent extreme inundation studies for Wellington, New Zealand, were executed using the numerical tool, XBeach. Here, the two-dimensional physical dynamics associated with multiple small embayments and both reef and sandy beach substrates, require large, high spatial resolution numerical model extents, informed by a multi-source elevation surface. XBGPU, is a translation of key XBeach features into code that permits GPU-based acceleration. The present study presents a comparison between XBeach and XBGPU for the same numerical model configuration and extents. Three model resolutions were employed, ranging from a typical desktop CPU based XBeach model resolution, to the highest resolution model that will require High Performance Computing (HPC) scale resources. Two CPU HPC facilities were used and five GPUs to investigate the scalability of both XBeach and XBGPU. The latter ranged from desktop grade units to GPUs associated with professional computing facilities. XBeach scalability is investigated by mean of the speed-up ratio, the time saving ratio and the computational efficiency. These are in reference to the computational speed of a model running on one CPU core. The XBGPU speed-up ratio is presented as a function of the slowest GPU. Direct comparisons between XBeach and XBGPU were achieved by using computational capacity as a metric. The results indicate that even a desktop grade GPU can compete with the computational efficiency of HPC-scale CPU facilities. Small model resolutions presented inefficient scaling on both high-performance CPUs and GPUs while the high-resolution model presented near linear scalability for the high-performance GPUs. Nevertheless, HPCs can be efficient to solve large computational problems if enough CPUs are employed.

Compressed basis GMRES on high-performance graphics processing units

Krylov methods provide a fast and highly parallel numerical tool for the iterative solution of many large-scale sparse linear systems. To a large extent, the performance of practical realizations of these methods is constrained by the communication bandwidth in current computer architectures, motivating the investigation of sophisticated techniques to avoid, reduce, and/or hide the message-passing costs (in distributed platforms) and the memory accesses (in all architectures). This article leverages Ginkgo’s memory accessor in order to integrate a communication-reduction strategy into the (Krylov) GMRES solver that decouples the storage format (i.e., the data representation in memory) of the orthogonal basis from the arithmetic precision that is employed during the operations with that basis. Given that the execution time of the GMRES solver is largely determined by the memory accesses, the cost of the datatype transforms can be mostly hidden, resulting in the acceleration of the iterative step via a decrease in the volume of bits being retrieved from memory. Together with the special properties of the orthonormal basis (whose elements are all bounded by 1), this paves the road toward the aggressive customization of the storage format, which includes some floating-point as well as fixed-point formats with mild impact on the convergence of the iterative process. We develop a high-performance implementation of the “compressed basis GMRES” solver in the Ginkgo sparse linear algebra library using a large set of test problems from the SuiteSparse Matrix Collection. We demonstrate robustness and performance advantages on a modern NVIDIA V100 graphics processing unit (GPU) of up to 50% over the standard GMRES solver that stores all data in IEEE double-precision.

Ginkgo: A Modern Linear Operator Algebra Framework for High Performance Computing

In this article, we presentGinkgo, a modern C++ math library for scientific high performance computing. While classical linear algebra libraries act on matrix and vector objects,Ginkgo’s design principle abstracts all functionality as “linear operators,” motivating the notation of a “linear operator algebra library.”Ginkgo’s current focus is oriented toward providing sparse linear algebra functionality for high performance graphics processing unit (GPU) architectures, but given the library design, this focus can be easily extended to accommodate other algorithms and hardware architectures. We introduce this sophisticated software architecture that separates core algorithms from architecture-specific backends and provide details on extensibility and sustainability measures. We also demonstrateGinkgo’s usability by providing examples on how to use its functionality inside the MFEM and deal.ii finite element ecosystems. Finally, we offer a practical demonstration ofGinkgo’s high performance on state-of-the-art GPU architectures.

Floquet Time-Dependent Configuration Interaction for Modeling Ultrafast Electron Dynamics

Time-dependent electronic structure methods are a valuable tool for simulating spectroscopic experiments. Recent advances in time-dependent configuration interaction (TDCI) algorithms have made them an attractive means of modeling many-electron dynamics, particularly for cases where multireference effects are essential. Here we present an extension to TDCI, Floquet TDCI, where the electronic wave function is expanded in a basis of light-dressed determinants. Our approach is based on our high-performance graphics processing unit (GPU) accelerated implementation of complete active space configuration interaction (CASCI). Simulations of two-photon absorption demonstrate that Floquet TDCI is well-suited for modeling dynamics in intense, ultrashort laser pulses. Accurate results are obtained for pulse energies up to ∼4 × 10-4 J/cm2 per pulse in the most difficult case explored here. By simulation of a set of molecules under continuous wave coupling, we demonstrate the ability of Floquet to describe the entanglement of light and multiple molecules in a cavity (i.e., a cavity polariton). Excellent computational performance is observed: a 320 fs propagation of a large dye (C30N2H22) with a 2 as timestep and a large active space (10 electrons in 11 orbitals), including a monochromatic pulse with three photon states, was performed in 3 h 6 min on a single Tesla V100 GPU. Our Floquet TDCI algorithm scales linearly with the number of photon states and exponentially with the number of photon colors included in the calculation. We argue that its energy-conserving nature makes Floquet TDCI well-suited to drive nonadiabatic molecular dynamics simulations.

EAIS: Energy-aware adaptive scheduling for CNN inference on high-performance GPUs

Recently, a large number of convolutional neural network (CNN) inference services have emerged on high-performance Graphic Processing Units (GPUs). However, GPUs are high power consumption units, and the energy consumption increases sharply along with the deployment of deep learning tasks. Although previous studies have considered the latency Service-Level-Objective (SLO) of inference services, they fail to directly take account of the energy consumption. Our investigation shows that coordinating batching and dynamic voltage frequency scaling (DVFS) settings can decrease the energy consumption of CNN inference. But it is affected by (i) larger configuration spaces; (ii) GPUs’ underutilization while data are transferred between CPUs and GPUs; (iii) fluctuating workloads. In this paper, we propose EAIS, an energy-aware adaptive scheduling framework that is comprised of a performance model, an asynchronous execution strategy, and an energy-aware scheduler. The performance model provides valid information about the performance characteristics of CNN inference services to shrink the feasible configuration space. The asynchronous execution strategy overlaps data upload and GPU execution to improve the system processing capacity. The energy-aware scheduler adaptively coordinates batching and DVFS according to fluctuating workloads to minimize energy consumption while meeting latency SLO. Our experimental results on NVIDIA Tesla M40 and V100 GPUs show that, compared to the state-of-the-art methods, EAIS decreases the energy consumption by up to 28.02% and improves the system processing capacity by up to 7.22% while meeting latency SLO. Besides, EAIS has been proved to have good versatility under different latency SLO constraints.

Artificial intelligence: machine learning for chemical sciences.

Research in molecular sciences witnessed the rise and fall of Artificial Intelligence (AI)/ Machine Learning (ML) methods, especially artificial neural networks, few decades ago. However, we see a major resurgence in the use of modern ML methods in scientific research during the last few years. These methods have had phenomenal success in the areas of computer vision, speech recognition, natural language processing (NLP), etc. This has inspired chemists and biologists to apply these algorithms to problems in natural sciences. Availability of high performance Graphics Processing Unit (GPU) accelerators, large datasets, new algorithms, and libraries has enabled this surge. ML algorithms have successfully been applied to various domains in molecular sciences by providing much faster and sometimes more accurate solutions compared to traditional methods like Quantum Mechanical (QM) calculations, Density Functional Theory (DFT) or Molecular Mechanics (MM) based methods, etc. Some of the areas where the potential of ML methods are shown to be effective are in drug design, prediction of high–level quantum mechanical energies, molecular design, molecular dynamics materials, and retrosynthesis of organic compounds, etc. This article intends to conceptually introduce various modern ML methods and their relevance and applications in computational natural sciences.Graphical abstract Synopsis Recent surge in the application of machine learning (ML) methods in fundamental sciences has led to a perspective that these methods may become important tools in chemical science. This perspective provides an overview of the modern ML methods and their successful applications in chemistry during the last few years.

Development of a scalable tabletop display using projection-based light field technology

We present a light-field display optical system design, state-of-the-art hardware technology, and computational software that are readily scalable (due to their modular structure) and provide naturally immersive and volumetric display systems. We explain the integration of 72 microprojectors into the tabletop display system. In addition, we used 6 workstations for all images and 18 high-performance graphics processing units. Sophisticated image generation through author tooling and the pixel re-arrangement algorithm in the Unity engine enabled us to produce complete three-dimensional images from the radiant rays of two-dimensional pixels using diffuser screens and microlens arrays.

Crack Detection on Concrete Surfaces Using Deep Encoder-Decoder Convolutional Neural Network: A Comparison Study Between U-Net and DeepLabV3+

Maintenance of infrastructures is a crucial activity to ensure safety using crack detection methods on concrete structures. However, most practice of crack detection is carried out manually, which is unsafe, highly subjective, and time-consuming. Therefore, a more accurate and efficient system needs to be implemented using artificial intelligence. Convolutional neural network (CNN), a subset of artificial intelligence, is used to detect cracks on concrete surfaces through semantic image segmentation. The purpose of this research is to compare the effectiveness of cutting-edge encoder-decoder architectures in detecting cracks on concrete surfaces using U-Net and DeepLabV3+ architectures with potential in biomedical, and sparse multiscale image segmentations, respectively. Neural networks were trained using cloud computing with a high-performance Graphics Processing Unit NVIDIA Tesla V100 and 27.4 GB of RAM. This study used internal and external data. Internal data consisted of simple cracks and were used as the training and validation data. Meanwhile, external data consisted of more complex cracks, which were used for further testing. Both architectures were compared based on four evaluation metrics in terms of accuracy, F1, precision, and recall. U-Net achieved segmentation accuracy = 96.57%, F1 = 87.55%, precision = 88.15%, and recall = 88.94%, while DeepLabV3+ achieved segmentation accuracy = 96.47%, F1 = 85.29%, precision = 92.07%, and recall = 81.84%. Experiment results (internal and external data) indicated that both architectures were accurate and effective in segmenting cracks. Additionally, U-Net and DeepLabV3+ exceeded the performance of previously tested architecture, namely FCN.

Game Effect Based on Particle System in Unity 3D

The gaming industry has been one of the main driving forces behind the development of advanced modern hardware such as multi-core, hyperthreaded processors, high-performance graphics processing units (GPUs), advanced sound processing devices, and extraordinary human-computer interface devices such as virtual reality helmets and brain sensor caps . The computer gaming industry has grown by leaps and bounds, becoming a mainstream software development sector, and earning billions of dollars in revenue each year. Computer games are like a book, a movie, or a museum .

New Two-Level L1 Data Cache Bypassing Technique for High Performance GPUs

On-chip caches of graphics processing units (GPUs) have contributed to improved GPU performance by reducing long memory access latency. However, cache efficiency remains low despite the facts that recent GPUs have considerably mitigated the bottleneck problem of L1 data cache. Although the cache miss rate is a reasonable metric for cache efficiency, it is not necessarily proportional to GPU performance. In this study, we introduce a second key determinant to overcome the problem of predicting the performance gains from L1 data cache based on the assumption that miss rate only is not accurate. The proposed technique estimates the benefits of the cache by measuring the balance between cache efficiency and throughput. The throughput of the cache is predicted based on the warp occupancy information in the warp pool. Then, the warp occupancy is used for a second bypass phase when workloads show an ambiguous miss rate. In our proposed architecture, the L1 data cache is turned off for a long period when the warp occupancy is not high. Our two-level bypassing technique can be applied to recent GPU models and improves the performance by 6% on average compared to the architecture without bypassing. Moreover, it outperforms the conventional bottleneck-based bypassing techniques.

Efficient Depth Map Creation with a Lightweight Deep Neural Network

Finding depth information with stereo matching using a deep learning algorithm for embedded systems has recently gained significant attention owing to emerging high-performance mobile graphics processing units (GPUs). Several researchers have proposed feasible small-scale CNNs that can run on a local GPU, but they still suffer from low accuracy and/or high computational requirements. In the method proposed in this study, pooling layers with padding and an asymmetric convolution filter are used to reduce computational costs and simultaneously maintain the accuracy of disparity. The patch size and number of layers are adjusted by analyzing the feature and activation maps. The proposed method forms a small-scale network algorithm suitable for a vision system at the edge and still exhibits high-disparity accuracy and low computational loads as compared to existing stereo-matching networks.

Моделирование газовой динамики гиперзвуковых летательных аппаратов с использованием модели высокотемпературного воздуха и графических процессоров

Проводится численное моделирование обтекания гиперзвукового летательного аппарата с использованием модели высокотемпературного воздуха и гибридной архитектуры на основе высокопроизводительных графических процессорных устройств. Расчеты проводятся на основе уравнений Эйлера, для дискретизации которых применяется метод конечных объемов на неструктурированных сетках. Приводятся результаты исследования эффективности расчета гиперзвуковых течений газа на графических процессорах. Обсуждается время счета, достигнутое при использовании моделей совершенного и реального газа. Numerical simulation of the flow around a hypersonic aircraft is carried out using a high-temperature air model and a hybrid architecture based on high-performance graphics processing units. The calculations are performed with the Euler equations discretized by the finite volume method on unstructured meshes. The scalability of the developed implementations of the model is studied and the results of the study of the efficiency of calculating hypersonic gas flows on graphics processors are analyzed. The computational time spent with the perfect and real gas models is discussed.

On deep learning techniques to boost monocular depth estimation for autonomous navigation

Inferring the depth of images is a fundamental inverse problem within the field of Computer Vision since depth information is obtained through 2D images, which can be generated from infinite possibilities of observed real scenes. Benefiting from the progress of Convolutional Neural Networks (CNNs) to explore structural features and spatial image information, Single Image Depth Estimation (SIDE) is often highlighted in scopes of scientific and technological innovation, as this concept provides advantages related to its low implementation cost and robustness to environmental conditions. In the context of autonomous vehicles, state-of-the-art CNNs optimize the SIDE task by producing high-quality depth maps, which are essential during the autonomous navigation process in different locations. However, such networks are usually supervised by sparse and noisy depth data, from Light Detection and Ranging (LiDAR) laser scans, and are carried out at high computational cost, requiring high-performance Graphic Processing Units (GPUs). Therefore, we propose a new lightweight and fast supervised CNN architecture combined with novel feature extraction models which are designed for real-world autonomous navigation. We also introduce an efficient surface normals module, jointly with a simple geometric 2.5D loss function, to solve SIDE problems. We also innovate by incorporating multiple Deep Learning techniques, such as the use of densification algorithms and additional semantic, surface normals and depth information to train our framework. The method introduced in this work focuses on robotic applications in indoor and outdoor environments and its results are evaluated on the competitive and publicly available NYU Depth V2 and KITTI Depth datasets.