• IEMEK Journal of Embedded Systems and Applications
• /
• v.11 no.2
• /
• pp.57-65
• /
• 2016

In this paper, we propose a system for efficient use of shared memory between CPU and GPU. The system, called Fusion Architecture, assures consistency of the shared memory and minimizes cache misses that frequently occurs on Heterogeneous System Architecture or Unified Virtual Memory based systems. It also maximizes the performance for memory intensive jobs by efficient allocation of GPU cores. To test between architectures on various scenarios, we introduce the Fusion Architecture Analyzer, which compares OpenMP, OpenCL, CUDA, and the proposed architecture in terms of memory overhead and process time. As a result, Proposed fusion architectures show that the Fusion Architecture runs benchmarks 55% faster and reduces memory overheads by 220% in average.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2012.11a
• /
• pp.52-53
• /
• 2012

Belief propagation (BP) is a commonly used global energy minimization algorithm for solving stereo matching problem in 3D reconstruction. However, it requires large memory bandwidth and data size. In this paper, we propose a novel memory-efficient algorithm of BP in stereo matching on the Graphics Processing Units (GPU). The data size and transfer bandwidth are significantly reduced by storing only a part of the whole message. In order to maintain the accuracy of the matching result, the local messages are reconstructed using shared memory available in GPU. Experimental result shows that there is almost an order of reduction in the global memory consumption, and 21 to 46% saving in memory bandwidth when compared to the conventional algorithm. The implementation result on a recent GPU shows that we can obtain 22.8 times speedup in execution time compared to the execution on CPU.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.4
• /
• pp.89-96
• /
• 2013

The graphic processor unit (GPU) has been developed to process not only graphic data but also general system data. It shows a better performance than CPU in algorithm for 3D graphics and parallel program. In order to execute algorithm for CPU on GPU, we should understand about GPU architectures and rewrite program considering parallel processing capability and new memory model of GPU. For this reasons, a performance prediction model for the algorithm and its predicted performance through GPU system are required. These can predict problems in GPU application development or construct a performance evaluation standard for GPU. In this paper, we applied the AES encryption algorithms on our performance model and accomplished performance prediction with high accuracy under a heavy workload.

• Journal of Broadcast Engineering
• /
• v.18 no.2
• /
• pp.204-214
• /
• 2013

Recently various research have been conducted relating to the implementation of signal processing or communication system by software using the massively parallel processing capability of the GPU. In this work, we focus on reducing software simulation time of 2K/8K FFT in DVB-T by using GPU. we estimate the processing time of the DVB-T system, which is one of the standards for DTV transmission, by CPU. Then we implement the FFT processing by the software using the NVIDIA's massively parallel GPU processor. In this paper we apply stream process method to reduce the overhead for data transfer between CPU and GPU, coalescing method to reduce the global memory access time and data structure design method to maximize the shared memory usage. The results show that our proposed method is approximately 20~30 times as fast as the CPU based FFT processor, and approximately 1.8 times as fast as the CUFFT library (version 2.1) which is provided by the NVIDIA when applied to the DVB-T 2K/8K mode FFT.

• Journal of KIISE
• /
• v.42 no.4
• /
• pp.522-529
• /
• 2015

Ever since the R-tree was proposed to index multi-dimensional data, many efforts have been made to improve its query performances. One common trend to improve query performance is to parallelize query processing with the use of multi-core architectures. To this end, a GPU-base R-tree has been recently proposed. However, even though a GPU-based R-tree can exhibit an improvement in query performance, it is limited in its ability to handle large volumes of data because GPUs have limited physical memory. To address this problem, we propose MGR-tree (Multi-GPU R-tree), which can manage large volumes of data by dividing nodes into multiple GPUs. Our experiments show that MGR-tree is up to 9.1 times faster than a sequential search on a GPU and up to 1.6 times faster than a conventional GPU-based R-tree.

• Journal of IKEEE
• /
• v.27 no.3
• /
• pp.265-272
• /
• 2023

In this paper, a GPU-accelerated kernel of range Doppler algorithm (RDA) was developed for real-time image formation based on frequency modulated continuous wave (FMCW) synthetic aperture radar (SAR). A pinned memory was used to minimize the data transfer time between the host and the GPU device, and the kernel was configured to perform all RDA operations on the GPU to minimize the number of data transfers. The dataset was obtained through the FMCW drone SAR experiment, and the GPU acceleration effect was measured in an intel i7-9700K CPU, 32GB RAM, and Nvidia RTX 3090 GPU environment. Including the data transfer time between host and devices, it was measured to be accelerated up to 3.41 times compared to the CPU, and when only the acceleration effect of operation was measured without including the data transfer time, it was confirmed that it could be accelerated up to 156 times.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.1
• /
• pp.111-116
• /
• 2019

Recently, GPU expands application domains from graphic processing to various kinds of parallel workloads. However, current GPU systems focus on the maximization of each workload's parallelism through simplified control rather than considering various workload characteristics. This paper classifies the resource usage characteristics of GPU workloads into computing-bound, memory-bound, and dependency-latency-bound, and quantifies the fine-grained bottleneck for efficient workload allocation. For example, we identify the exact bottleneck resources such as single function unit, double function unit, or special function unit even for the same computing-bound workloads. Our analysis implies that workloads can be allocated together if fine-grained bottleneck resources are different even for the same computing-bound workloads, which can eventually contribute to efficient workload allocation in GPU.

• KIISE Transactions on Computing Practices
• /
• v.22 no.7
• /
• pp.303-309
• /
• 2016

Recently, an increasing number of embedded processors such as ARM Mali begin to support GPGPU programming frameworks, such as OpenCL. Thus, GPGPU technologies that have been used in PC and server environments are beginning to be applied to the embedded systems. However, many embedded systems have different architectural characteristics compare to traditional PCs and low-power consumption and real-time performance are also important performance metrics in these systems. In this paper, we implement a parallel AES cryptographic algorithm for a modern embedded GPU using OpenCL, a standard parallel computing framework, and compare performance against various baselines. Experimental results show that the parallel GPU AES implementation can reduce the response time by about 1/150 and the energy consumption by approximately 1/290 compare to OpenMP implementation when 1000KB input data is applied. Furthermore, an additional 100 % performance improvement of the parallel AES algorithm was achieved by exploiting the characteristics of embedded GPUs such as removing copying data between GPU and host memory. Our results also demonstrate that higher performance improvement can be achieved with larger size of input data.

• Journal of KIISE:Databases
• /
• v.37 no.5
• /
• pp.275-284
• /
• 2010

GPUs are stream processors based on multi-cores, which can process large data with a high speed and a large memory bandwidth. Furthermore, GPUs are less expensive than multi-core CPUs. Recently, usage of GPUs in general purpose computing has been wide spread. The CUDA architecture from Nvidia is one of efforts to help developers use GPUs in their application domains. In this paper, we propose techniques to parallelize a skyline algorithm which uses a simple nested loop structure. In order to employ the CUDA programming model, we apply our optimization techniques to make our skyline algorithm fit into the performance restrictions of the CUDA architecture. According to our experimental results, we improve the original skyline algorithm by 80% with our optimization techniques.

• Journal of Korea Multimedia Society
• /
• v.14 no.10
• /
• pp.1229-1237
• /
• 2011

The volume ray-casting method is one of the direct volume rendering methods that produces high-quality images as well as manipulates semi-transparent object. Although the volume ray-casting method produces high-quality image by sampling in the region of interest, its rendering speed is slow since the color acquisition process is complicated for repetitive memory reference and accumulation of sample values. Recently, the GPU-based acceleration techniques are introduced. However, they require pre-processing or additional memory. In this paper, we propose efficient point-primitive based method to overcome complicated computation of GPU ray-casting. It presents semi-transparent objects, however it does not require preprocessing and additional memory. Our method is fast since it generates point-primitives from volume dataset during sampling process and it projects the primitives onto the image plane. Also, our method can easily cope with OTF change because we can add or delete point-primitive in real-time.