NVIDIA Compute Unified Device Architecture Research Articles

Comparative study of CUDA-based parallel programming in C and Python for GPU acceleration of the 4th order Runge-Kutta method

In this paper, a comparative study is presented on the application of General-purpose Computing on Graphics Processing Units for solving the point reactor kinetics equations through the utilization of the 4th Order Runge-Kutta (RK4) method using the programming languages C and Python. Sequential and parallel algorithms of the RK4 method were developed in C/C++ and Python, with parallel algorithms specifically designed to operate on Graphics Processing Units (GPUs) utilizing the NVIDIA Compute Unified Device Architecture (CUDA) as the programming platform. As an experiment, the execution time for the sequential and parallel algorithms were compared for a reactivity value of ρ = 0.003 and a simulation time of t = 100 s. The parallel C and Python algorithms achieved, respectively, speedups of 9.33 and 409.7 when comparing the execution time on the best GPU utilized (RTX 3070Ti) with the best CPU (3600XT), while still maintaining numerical precision.

Study on the flooding characteristics of a deep-water submarine based on δ plus-smoothed particle hydrodynamics method and graphic processing units acceleration

The process of a damaged ship flooding is a complicated free surface flow problem. There is a complex coupling effect between the ship cabin and the flow inside and outside of the cabin. In this paper, a GPU (graphic processing unit)-δ+-SPH (smoothed particle hydrodynamics) numerical model for the cabin flooding in deep-water environments is developed based on GPU parallel acceleration technology and Nvidia's CUDA (compute unified device architecture). First, the computational accuracy and efficiency of this numerical model are verified by experiments results on the water flooding of a simple damaged cabin model. Furthermore, the flooding characteristics of a submarine cabin are analyzed, considering different numbers of damaged cabins, depths, and opening positions. Finally, the progressive flooding and the dynamic response characteristics of a full-scale submarine model are investigated. The results show that the process of progressive flooding in a submarine cabin is characterized by its rapidity and intensity. Different factors, for example, damaged cabin numbers, cabin depths, and opening positions, have great influences on the process of flooding and the motion of the submarine cabin. This study can offer valuable technical assistance in the post-damage remediation process.

GPU Implementation of Graph-Regularized Sparse Unmixing With Superpixel Structures

To enhance spectral unmixing performance, a large number of algorithms have simultaneously investigated spatial and spectral information in hyperspectral images. However, sophisticated algorithms with high computational complexity can be very time-consuming when a large amount of data are involved in processing hyperspectral images. In this paper, we first introduce a group sparse graph-regularized unmixing method with superpixel structure, to promote piece-wise consistency of abundances and reduce computational burden. Segmenting the image into several non-overlapped superpixels also enables to decompose the unmixing problem into uncoupled subproblems that can be processed in parallel. An implementation for the proposed algorithm on graphics processing units (GPUs) is then developed based on the NVIDIA Compute Unified Device Architecture (CUDA) framework. The proposed scheme achieves parallelism at both the intra-superpixel and inter-superpixel levels, where multiple concurrent streams have been used to enable multiple kernels to execute on the device simultaneously. Simulation results with a series of experiments demonstrate advantages of the proposed algorithm. The performance of the GPU implementation also illustrates that parallel scheme largely expedites the implementation.

Software Package for Remote Diagnostics of Agricultural Machinery Condition

Artificial intelligence is stated to be more and more widely used in agriculture, as well as for the diagnostics of the agricultural machinery condition. It was noted that in besides software, new computing devices are developed that enable processing and storing large amounts of data. (Research purpose) To create a neural network-based software package for remote diagnostics of the limit state of machinery individual components and assemblies. (Materials and methods) Foreign studies within the problem area were analysed. It was found out that for data collection for artificial intelligence there exist STM32 and Arduino microcontroller-based devices, and the Nvidia CUDA (Compute Unified Device Architecture) hardware and software platform is used. For the software was developed in the C / C ++ programming language, and the MS SQL Server database were used as a repository. The general software is emphasized to be able to run on all major operating systems such as Windows, Mac OS, Linux. The role of neural network is argued to be important since it integrated all program blocks and provides its own analysis. (Results and discussion) The information from the diagnostics devices is accumulated in a database. The neural network created on the basis of this database is constantly learning and simultaneously analyzing incoming data in real time, automatically issuing its recommendations. It was found that the neural network created by the employees of the Federal Scientific Agroengineering Center VIM has more functional options, for example, it is able to work directly with devices and conduct a more detailed technical analysis. (Conclusions) A neural network for equipment condition diagnostics was created, which increases the efficiency of decision-making in case of repair, and improves forecast and predictability. The criteria for equipment operation were proposed.

PERCEPTRON: an open-source GPU-accelerated proteoform identification pipeline for top-down proteomics.

PERCEPTRON is a next-generation freely available web-based proteoform identification and characterization platform for top-down proteomics (TDP). PERCEPTRON search pipeline brings together algorithms for (i) intact protein mass tuning, (ii) de novo sequence tags-based filtering, (iii) characterization of terminal as well as post-translational modifications, (iv) identification of truncated proteoforms, (v) in silico spectral comparison, and (vi) weight-based candidate protein scoring. High-throughput performance is achieved through the execution of optimized code via multiple threads in parallel, on graphics processing units (GPUs) using NVidia Compute Unified Device Architecture (CUDA) framework. An intuitive graphical web interface allows for setting up of search parameters as well as for visualization of results. The accuracy and performance of the tool have been validated on several TDP datasets and against available TDP software. Specifically, results obtained from searching two published TDP datasets demonstrate that PERCEPTRON outperforms all other tools by up to 135% in terms of reported proteins and 10-fold in terms of runtime. In conclusion, the proposed tool significantly enhances the state-of-the-art in TDP search software and is publicly available at https://perceptron.lums.edu.pk. Users can also create in-house deployments of the tool by building code available on the GitHub repository (http://github.com/BIRL/Perceptron).

Parallel implementation of a numerical method for solving transport equations for the mesoscale meteorological model TSUNM3

A parallel algorithm for numerical solution of a generalized unsteady 3-dimensional advection-diffusion equation for a mesoscale meteorological model TSUNM3 is considered. Efficiency of parallel implementation of the solver with Message Passing Interface (MPI), Open Multi-Processing (OpenMP), Open Accelerators (OpenACC), and NVidia Compute Unified Device Architecture (CUDA) technologies was compared. The efficiency of the MPI-parallelized code of TSUNM3 model with a two-dimensional domain decomposition strategy applied is not less than 75% on 256–512 processes on Cyberia supercomputer of Tomsk State University.

Real Time Face Expression Recognition along with Balanced FER2013 Dataset using CycleGAN

Human face expression recognition is an active research area that has massive applications in medical field, crime investigation, marketing, online learning, automobile safety and video games. The first part of this research defines a deep neural network model-based framework for recognizing the seven main types of facial expression, which are found in all cultures. The proposed methodology involves four stages: (a) pre-processing the FER2013 dataset through relabeling to avoid misleading results and getting rid of non-face and non-frontal faces; (b) design of an efficient stable Cycle Generative Adversarial Network (CycleGAN), which provides unsupervised expression-to-expression translation. The CycleGAN has been designed and trained with a new cycle consistency loss. (c) Generating new images to overcome the class imbalance and finally (d) building the DNN architecture for recognizing the face sign expression, using the pretrained VGG-Face model with vggface weights. The second part encompasses the design of a GPU-accelerated face expression recognition system for real time video sequences using NVIDIAs Compute Unified Device Architecture (CUDA). OpenCV library has been compiled from scratch with CUDA and NVIDIA CUDA Deep Neural Network library, cuDNN. For face detection stage Haar Cascaded and deep learning were used and tested using both CPU and GPU as backend. Results show that the designed model run time to recognize a face sign is 0.44 seconds. Besides, the average test accuracy has been increased from 64% for the original FER2013 dataset to 91.76% for the modified balanced version using the same transfer learning model.

Parallel implementation of the non-overlapping template matching test using CUDA

NIST (National Institute of Standards and Technology) statistical test recognized as the most authoritative is widely used in verifying the randomness of binary sequences. The Non-overlapping Template Matching Test as the 7th test of the NIST Test Suit is remarkably time consuming and the slow performance is one of the major hurdles in the testing process. In this paper, we present an efficient bit-parallel matching algorithm and segmented scan-based strategy for execution on Graphics Processing Unit (GPU) using NVIDIA Compute Unified Device Architecture (CUDA). Experimental results show the significant performance improvement of the parallelized Non-overlapping Template Matching Test, the running speed is 483 times faster than the original NIST implementation without attenuating the test result accuracy.

Fingerprint Matching Using Bozorth3 Algorithm and Parallel Computation on NVIDIA Compute Unified Device Architecture

This paper studied fingerprint matching employing Bozorth3 Algorithm for matching fingerprint and parallel computation employing NVIDIA Compute Unified Device Architecture (NVIDIA CUDA). The objective of this study obtains the percentage and time processing of matching fingerprints. In this study, the fingerprint matching is done with parallel computing is applied to the GPU (Graphics Processing Unit). GPU device used in this study is the CUDA (Compute Unified Device Architecture), which is an Application Programming Interface (API) developed by NVIDIA. The development of applications with fingerprint matching serial computing on CPU and parallel computing on GPU can be applied to the CUDA API. The results from this study can be found in the performance process on the CPU and GPU. The results of this research are the process on CUDA execution time is better than the execution time on the CPU, the process is done at both the computation is to find a match in the fingerprint value.

Real-Time 3-D Imaging Using an Air-Coupled Ultrasonic Phased-Array.

We present an air-coupled ultrasonic imaging system based on a 40-kHz 8×8 phased-array for 3-D real-time localization of multiple objects in the far-field. By attaching a waveguide to the array, the effective interelement spacing is reduced to half wavelength. This enables grating lobe-free transmit and receive beamforming with a uniform rectangular array of efficient low-cost transducers. The system further includes custom transceiver electronics, an field programmable gate array (FPGA) system-on-chip and a PC for GPU accelerated frequency domain signal processing, consisting of matched filtering, conventional beamforming, and envelope extraction using Nvidia Compute Unified Device Architecture (CUDA) and OpenGL for visualization. The uniform rectangular layout allows utilizing multiple transmit and receive methods, known from medical imaging applications. Thus, the system is dynamically adaptable to maximize the frame rate or detection range. One implemented method demonstrates the real-time capability by transmitting a hemispherical pulse (HP) with a single transducer to irradiate the surroundings simultaneously, whereas all transducers are used for echo reception. The imaging properties, such as axial and lateral resolution, field of view and range of view, are characterized in an anechoic chamber. The object localization is validated for a horizontal and vertical field of view of ±80° and a range of view of 0.5-3 m with 29 frames/s. Using the same system, a comparison between the HP method and the dynamic transmit beamforming method, which transmits multiple sequential beamformed pulses for long-range localization, is provided.

GPU-accelerated Lagrangian heuristic for multidimensional assignment problems with decomposable costs

In this paper, we describe a GPU-accelerated parallel algorithm for the axial Multidimensional Assignment Problem with Decomposable Costs (MDADC), which is one of the most fundamental formulations for data association. MDADC is known to be NP-hard and is large-dimensioned in most realistic cases; hence, heuristic solutions with qualified optimality gaps is the best one can hope for, given the state-of-knowledge. The main contribution of this paper is an efficient parallelization of the Lagrangian subgradient search algorithm specifically targeted towards the Graphics Processing Units (GPUs) based on the NVIDIA Compute Unified Device Architecture (CUDA). A GPU-accelerated Linear Assignment Problem (LAP) solver is leveraged in concert with the Lagrangian scheme for further speed-up. We also implemented a multi-GPU variant of this algorithm which maintains a good speedup profile, when tested on problems with 31 billion variables, on up to 128 GPUs.

Accelerating a Geometrical Approximated PCA Algorithm Using AVX2 and CUDA

Remote sensing data has known an explosive growth in the past decade. This has led to the need for efficient dimensionality reduction techniques, mathematical procedures that transform the high-dimensional data into a meaningful, reduced representation. Projection Pursuit (PP) based algorithms were shown to be efficient solutions for performing dimensionality reduction on large datasets by searching low-dimensional projections of the data where meaningful structures are exposed. However, PP faces computational difficulties in dealing with very large datasets—which are common in hyperspectral imaging, thus raising the challenge for implementing such algorithms using the latest High Performance Computing approaches. In this paper, a PP-based geometrical approximated Principal Component Analysis algorithm (gaPCA) for hyperspectral image analysis is implemented and assessed on multi-core Central Processing Units (CPUs), Graphics Processing Units (GPUs) and multi-core CPUs using Single Instruction, Multiple Data (SIMD) AVX2 (Advanced Vector eXtensions) intrinsics, which provide significant improvements in performance and energy usage over the single-core implementation. Thus, this paper presents a cross-platform and cross-language perspective, having several implementations of the gaPCA algorithm in Matlab, Python, C++ and GPU implementations based on NVIDIA Compute Unified Device Architecture (CUDA). The evaluation of the proposed solutions is performed with respect to the execution time and energy consumption. The experimental evaluation has shown not only the advantage of using CUDA programming in implementing the gaPCA algorithm on a GPU in terms of performance and energy consumption, but also significant benefits in implementing it on the multi-core CPU using AVX2 intrinsics.

An Efficient Acceleration of Solving Heat and Mass Transfer Equations with the Second Kind Boundary Conditions in Capillary Porous Radially Composite Cylinder Using Programmable Graphics Hardware

With the recent developments in computing technology, increased efforts have gone into the simulation of various scientific methods and phenomenon in engineering fields. One such case is the simulation of heat and mass transfer equations which is becoming more and more important in analyzing various scenarios in engineering applications. Analysing the heat and mass transfer phenomenon under various environmental conditions require us to simulate it. However, this process of numerical solution of heat and mass transfer equations is very time consuming. Therefore, this paper aims at utilizing one of the acceleration techniques developed in the graphics community that exploits a graphics processing unit (GPU) which is applied to the numerical solutions of heat and mass transfer equations. The nVidia Compute Unified Device Architecture (CUDA) programming model can be a good method of applying parallel computing to program the graphical processing unit. This paper shows a good improvement in the performance, while solving the heat and mass transfer equations for a capillary porous radially composite cylinder with the second kind of boundary conditions, numerically running on GPU. This heat and mass transfer simulation is implemented using CUDA platform on nVidia Quadro FX 4800 graphics card. Our experimental results depict the drastic performance improvement when GPU is used to perform heat and mass transfer simulation. GPU can significantly accelerate the performance with a maximum observed speedup of more than 8 fold times. Therefore, the GPU is a good approach to accelerate the heat and mass transfer simulation.

Parallel implementation of a numerical method for solving a three-dimensional transport equation for a mesoscale meteorological model

Abstract A parallel algorithm for numerical solution of a generalized unsteady 3-dimensional advection-diffusion equation for a mesoscale meteorological model is considered. Efficiency of parallel implementations of the solver with Message Passing Interface (MPI), Open Multi-Processing (OpenMP), and NVidia Compute Unified Device Architecture (CUDA) technologies was compared. It was shown that on the mesh of 25625632 nodes the speedup of the program reaches 18 for OpenMP and MPI implementations and 38 for CUDA.

Fast differential box-counting algorithm on GPU

The differential box-counting (DBC) algorithm is the most widely used method for calculating the fractal dimension (FD) of grayscale images. FD analysis of grayscale images has important applications in fields such as shape classification, texture analysis and image segmentation. Nowadays these kinds of images can reach a very large size, especially in 3D, and there is a trend of increasing dataset size; this means that a current dataset of this type of image may need high computation times for computing the DBC. In this paper, we present an efficient implementation on graphics processing unit (GPU) of the original DBC algorithm in its optimized version for parallel processing. Our implementation with NVIDIA Compute Unified Device Architecture (CUDA) computes the DBC with high branch efficiency and very low GPU serialization. This fact allows us to obtain a very efficient GPU implementation of the algorithm. We tested our implementation on two different hardware/software platforms for a set of 2D and 3D grayscale images of increasing size. The results showed that this GPU implementation scaled very well and achieved a speedup of up to 52× with respect to a CPU single-thread implementation of the same algorithm. Against an OpenACC multi-thread implementation in CPU, our CUDA algorithm obtains a speedup of up to 6×.

Parallelizing Multiple Flow Accumulation Algorithm using CUDA and OpenACC

Watershed analysis, as a fundamental component of digital terrain analysis, is based on the Digital Elevation Model (DEM), which is a grid (raster) model of the Earth surface and topography. Watershed analysis consists of computationally and data intensive computing algorithms that need to be implemented by leveraging parallel and high-performance computing methods and techniques. In this paper, the Multiple Flow Direction (MFD) algorithm for watershed analysis is implemented and evaluated on multi-core Central Processing Units (CPU) and many-core Graphics Processing Units (GPU), which provides significant improvements in performance and energy usage. The implementation is based on NVIDIA CUDA (Compute Unified Device Architecture) implementation for GPU, as well as on OpenACC (Open ACCelerators), a parallel programming model, and a standard for parallel computing. Both phases of the MFD algorithm (i) iterative DEM preprocessing and (ii) iterative MFD algorithm, are parallelized and run over multi-core CPU and GPU. The evaluation of the proposed solutions is performed with respect to the execution time, energy consumption, and programming effort for algorithm parallelization for different sizes of input data. An experimental evaluation has shown not only the advantage of using OpenACC programming over CUDA programming in implementing the watershed analysis on a GPU in terms of performance, energy consumption, and programming effort, but also significant benefits in implementing it on the multi-core CPU.

Fast and robust flow simulations in discrete fracture networks with GPGPUs

A new approach for flow simulation in very complex discrete fracture networks based on PDE-constrained optimization has been recently proposed in Berrone et al. (SIAM J Sci Comput 35(2):B487–B510, 2013b; J Comput Phys 256:838–853, 2014) with the aim of improving robustness with respect to geometrical complexities. This is an essential issue, in particular for applications requiring simulations on geometries automatically generated like the ones used for uncertainty quantification analyses and hydro-mechanical simulations. In this paper, implementation of this approach in order to exploit Nvidia Compute Unified Device Architecture is discussed with the main focus to speed up the linear algebra operations required by the approach, being this task the most computational demanding part of the approach. Furthermore, two different approaches for linear algebra operations and two storage formats for sparse matrices are compared in terms of computational efficiency and memory constraints.

Brain abnormality detection using template matching

Abstract Magnetic resonance imaging (MRI) is a widely used imaging modality to evaluate brain disorders. MRI generates huge volumes of data, which consist of a sequence of scans taken at different instances of time. As the presence of brain disorders has to be evaluated on all magnetic resonance (MR) sequences, manual brain disorder detection becomes a tedious process and is prone to inter- and intra-rater errors. A technique for detecting abnormalities in brain MRI using template matching is proposed. Bias filed correction is performed on volumetric scans using N4ITK filter, followed by volumetric registration. Normalized cross-correlation template matching is used for image registration taking into account, the rotation and scaling operations. A template of abnormality is selected which is then matched in the volumetric scans, if found, the corresponding image is retrieved. Post-processing of the retrieved images is performed by the thresholding operation; the coordinates and area of the abnormality are reported. The experiments are carried out on the glioma dataset obtained from Brain Tumor Segmentation Challenge 2013 database (BRATS 2013). Glioma dataset consisted of MR scans of 30 real glioma patients and 50 simulated glioma patients. NVIDIA Compute Unified Device Architecture framework is employed in this paper, and it is found that the detection speed using graphics processing unit is almost four times faster than using only central processing unit. The average Dice and Jaccard coefficients for a wide range of trials are found to be 0.91 and 0.83, respectively.

Performance Evaluation of SW Algorithm on NVIDIA GeForce GTX TITAN X Graphic Processing Unit (GPU)

Nowadays, the requirement for high performance and sensitive alignment tools have increased after the advantage of the Deoxyribonucleic Acid (DNA) and molecular biology has been figured out through Bioinformatics study. Therefore, this paper reports the performance evaluation of parallel Smith-Waterman Algorithm implementation on the new NVIDIA GeForce GTX Titan X Graphic Processing Unit (GPU) compared to the Central Processing Unit (CPU) running on Intel® CoreTM i5-4440S CPU 2.80GHz. Both of the design were developed using C-programming language and targeted to the respective platform. The code for GPU was developed and compiled using NVIDIA Compute Unified Device Architecture (CUDA). It clearly recorded that, the performance of GPU based computational is better compared to the CPU based. These results indicate that the GPU based DNA sequence alignment has a better speed in accelerating the computational process of DNA sequence alignment.

An Accelerated Tool for Flood Modelling Based on Iber

This paper presents Iber+, a new parallel code based on the numerical model Iber for two-dimensional (2D) flood inundation modelling. The new implementation, which is coded in C++ and takes advantage of the parallelization functionalities both on CPUs (central processing units) and GPUs (graphics processing units), was validated using different benchmark cases and compared, in terms of numerical output and computational efficiency, with other well-known hydraulic software packages. Depending on the complexity of the specific test case, the new parallel implementation can achieve speedups up to two orders of magnitude when compared with the standard version. The speedup is especially remarkable for the GPU parallelization that uses Nvidia CUDA (compute unified device architecture). The efficiency is as good as the one provided by some of the most popular hydraulic models. We also present the application of Iber+ to model an extreme flash flood that took place in the Spanish Pyrenees in October 2012. The new implementation was used to simulate 24 h of real time in roughly eight minutes of computing time, while the standard version needed more than 15 h. This huge improvement in computational efficiency opens up the possibility of using the code for real-time forecasting of flood events in early-warning systems, in order to help decision making under hazardous events that need a fast intervention to deploy countermeasures.