• ETRI Journal
• /
• v.45 no.6
• /
• pp.1035-1045
• /
• 2023

We propose a novel graphics processing unit (GPU) algorithm that can handle a large-scale 3D fast Fourier transform (i.e., 3D-FFT) problem whose data size is larger than the GPU's memory. A 1D FFT-based 3D-FFT computational approach is used to solve the limited device memory issue. Moreover, to reduce the communication overhead between the CPU and GPU, we propose a 3D data-transposition method that converts the target 1D vector into a contiguous memory layout and improves data transfer efficiency. The transposed data are communicated between the host and device memories efficiently through the pinned buffer and multiple streams. We apply our method to various large-scale benchmarks and compare its performance with the state-of-the-art multicore CPU FFT library (i.e., fastest Fourier transform in the West [FFTW]) and a prior GPU-based 3D-FFT algorithm. Our method achieves a higher performance (up to 2.89 times) than FFTW; it yields more performance gaps as the data size increases. The performance of the prior GPU algorithm decreases considerably in massive-scale problems, whereas our method's performance is stable.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.10
• /
• pp.135-141
• /
• 2013

In this paper, we installed three high brightness red, green, and blue LED in one socket and made one pixel unit. And we also developed the full-color display board module and control unit which can express various images such as text, graphics, video image with the combination of pixel units and a number of modules. LED display driver module have a driver circuit within the combination of the RGB pixel dot on unit area. These modules of the existing form can be high priced because of implementation a fixed resolution in specific space and installation space. To overcome these shortcomings, we developed a LED driver and LED pixel modules free in array at random pitch intervals. Display board module of this paper enabled to display smoothly video image which have many data processing quantity through dragging data speed up 36 frames per second. Also there are an effect which is provided more clear image because of improving the flickering of the existing display board.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.2
• /
• pp.270-276
• /
• 2018

Apache Spark is one of the high performance in-memory computing frameworks for big-data processing. Recently, to improve the performance, general-purpose computing on graphics processing unit(GPGPU) is adapted to Apache Spark framework. Previous Spark-GPGPU frameworks focus on overcoming the difficulty of an implementation resulting from the difference between the computation environment of GPGPU and Spark framework. In this paper, we propose a Spark framework based on a heterogenous pipeline computing with OpenCL to further improve the performance. The proposed framework overlaps the Java-to-Native memory copies of CPU with CPU-GPU communications(DMA) and GPU kernel computations to hide the CPU idle time. Also, CPU-GPU communication buffers are implemented with switching dual buffers, which reduce the mapped memory region resulting in decreasing memory mapping overhead. Experimental results showed that the proposed Spark framework based on a heterogenous pipeline computing with OpenCL had up to 2.13 times faster than the previous Spark framework using OpenCL.

• Journal of Broadcast Engineering
• /
• v.17 no.6
• /
• pp.964-975
• /
• 2012

Recently the attention on digital hologram that is regarded as to be the final goal of the 3-dimensional video technology has been increased. A digital hologram can be generated with a depth and a RGB image. We proposed a new system to capture RGB and depth images and to convert them to digital holograms. First a new cold mirror was designed and produced. It has the different transmittance ratio against various wave length and can provide the same view and focal point to the cameras. After correcting various distortions with the camera system, the different resolution between depth and RGB images was adjusted. The interested object was extracted by using the depth information. Finally a digital hologram was generated with the computer generated hologram (CGH) algorithm. All algorithms were implemented with C/C++/CUDA and integrated in LabView environment. A hologram was calculated in the general-purpose computing on graphics processing unit (GPGPU) for high-speed operation. We identified that the visual quality of the hologram produced by the proposed system is better than the previous one.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.2
• /
• pp.11-19
• /
• 2020

Modern GPU can execute general purpose computation on the graphic processing unit, and provide high performance by exploiting many core on GPU. To run AES algorithm efficiently, parallel computational resources are required. However, computational resource of CPU architecture are not enough to cryptographic algorithm such as AES whereas GPU architecture has mass parallel computation resources. Therefore, this paper reduce the time to execute AES by employing parallel computational resource on GPGPU. Unfortunately, AES cannot utilize computational resource on GPGPU since it isn't suitable to GPGPU architecture. In this paper, IPC based dynamic SM management technique are proposed to efficiently execute AES on GPGPU. IPC based dynamic SM management can increase and decrease the number of active SMs by using IPC in run-time. According to simulation results, proposed technique improve the performance by increasing resource utilization compared to baseline GPGPU architecture. The results show that AES improve the performance by 41.2% on average.

• Journal of Satellite, Information and Communications
• /
• v.10 no.3
• /
• pp.94-101
• /
• 2015

The display device is an important video information communication system device to connect between human and device. it transfers the information as characters, shapes, images and pattern to enable recognizing by eyes. Theres absolutely needs some key functions and role to quickly display informations. It can analyse a information through a PPI Scope of a cathode-ray tube(CRT) displays information which can perform a role. this research proposed a radar device to display informations as received signal. The radar display researches can apply to fixed function graphics pipeline algorithms of the large capacity type through a vertical blanking interval and buffer swap of display unit. Also, it can be possible to apply to performed algorithms to FPGA logic without high-performance graphics processing unit GPU through synchronization which can implement a display system. In this paper, we improved the affordability and reliability through proposed research. 이So, we have studied the radar display unit which can change a flat display from radar display of CRT radar display.

• Journal of KIISE:Software and Applications
• /
• v.36 no.4
• /
• pp.253-260
• /
• 2009

Many algorithms for computer vision and pattern recognition have recently been implemented on GPU (graphic processing unit) for faster computational times. However, the implementation has two problems. First, the programmer should master the fundamentals of the graphics shading languages that require the prior knowledge on computer graphics. Second, in a job that needs much cooperation between CPU and GPU, which is usual in image processing and pattern recognition contrary to the graphic area, CPU should generate raw feature data for GPU processing as much as possible to effectively utilize GPU performance. This paper proposes more quick and efficient implementation of neural networks on both GPU and multi-core CPU. We use CUDA (compute unified device architecture) that can be easily programmed due to its simple C language-like style instead of GPU to solve the first problem. Moreover, OpenMP (Open Multi-Processing) is used to concurrently process multiple data with single instruction on multi-core CPU, which results in effectively utilizing the memories of GPU. In the experiments, we implemented neural networks-based text extraction system using the proposed architecture, and the computational times showed about 15 times faster than implementation on only GPU without OpenMP.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.436-442
• /
• 2009

Tomography images reconstructed from conebeam CT make it possible to observe inside of the projected object without any damage, and so it has been widely used in the industrial and medical fields. Recent advanced imaging equipment can produce high-resolution CT images. However, it takes much time to reconstruct the obtained large dataset. To reduce the time to reconstruct CT images, we propose an accelerating method using GPU (graphics processing unit). Reconstruction consists of mainly two parts, filtering and back-projection. In filtering phase, we applied 4ch image compression method and in back-projection phase, computation reduction method using depth test is applied. The experimental results show that the proposed method accelerates the speed 50 times than the CPU-based program optimized with OpenMP by utilizing the high-computing power of parallelized GPU.

• Journal of Biomedical Engineering Research
• /
• v.34 no.3
• /
• pp.135-140
• /
• 2013

We have developed an ultra wide-field of view Optical Coherence Tomography(OCT) which has capability to 2D and 3D views of cross-sectional structure of in vivo human retina. Conventional OCT has a limitation in visualizing the entire retina due to a reduced field of view. We designed an optical setup to significantly improve the lateral scanning range to be more than 20 mm. The entire human retinal structure in 2D and 3D was reported in this paper with the developed OCT system. Also, we empirically searched an optimized image size for real time visualization by analyzing variation of the frame rate with different lateral scan points. The size was concluded to be $1024{\times}2000{\times}300$ pixels which took 9 seconds for visualization.

• The Bulletin of The Korean Astronomical Society
• /
• v.39 no.2
• /
• pp.86.2-86.2
• /
• 2014

We present a new ray-tracing code, "Odyssey", for the Kerr spacetime accelerated by the Graphics Processing Unit (GPU). Taking advantage of the ability of nVidia graphic cards to evaluate trajectories of a large amount of photon simultaneously, the code is two orders of magnitude as fast as the previous CPU-based code corresponding to the speed of few nanoseconds per photon per time step. In the light of the Graphic User Interface (GUI) powered by the GPU-enhanced 2D/3D displaying technique, DirectX, it is feasible for users to manipulate diverse results such as rotating and zooming in/out the trajectories of photon instantly near the black hole. Thus the Odyssey can serve as a tool not only for scientific but also for the educational purpose. We discuss possible applications in detail in light of several results such as the shape of the silhouette of a black hole, the shape of a hot spot orbiting a black hole, and 3D photon trajectories.