• 전자통신동향분석
• /
• 제39권5호
• /
• pp.98-107
• /
• 2024

Applications such as artificial intelligence continue to grow in complexity and scale. Thus, the demand for scalable computing is increasing for achieving faster data processing and improved efficiency. This requirement has led to the development of memory-centric computing and high-speed fabric interconnection technologies. Memory-centric computing reduces the latency and enhances the system performance by shifting the focus from the central processing unit to the memory, whereas high-speed fabric interconnects enable efficient data transfer across various computing resources. Technologies such as Gen-Z, OpenCAPI, and CCIX have been integrated into the CXL (Compute Express Link) standard since 2019 to improve communication and cache coherence. Ethernet-based interconnects such as RoCE, InfiniBand, and OmniXtend also play a crucial role in providing high-speed data transfer and low latency. We explore the latest trends and prospects of these technologies, highlighting their benefits and applications.

• 인터넷정보학회논문지
• /
• 제22권5호
• /
• pp.57-67
• /
• 2021

Modern cloud computing is rapidly changing from traditional hypervisor-based virtual machines to container-based cloud-native environments. Due to limitations in I/O performance required for both virtual machines and containers, the use of high-speed storage (SSD, NVMe, etc.) is increasing, and in-memory computing using main memory is also emerging. Running a virtual environment on main memory gives better performance compared to other storage arrays. However, RAM used as main memory is expensive and due to its volatile characteristics, data is lost when the system goes down. Therefore, additional work is required to run the virtual environment in main memory. In this paper, we propose a hybrid in-memory storage that combines a block storage such as a high-speed SSD with main memory to safely operate virtual machines and containers on main memory. In addition, the proposed storage showed 6 times faster write speed and 42 times faster read operation compared to regular disks for virtual machines, and showed the average 12% improvement of container's performance tests.

• 한국정보통신학회논문지
• /
• 제24권8호
• /
• pp.985-992
• /
• 2020

엣지 컴퓨팅은 데이터를 처리하고 연산하는 곳이 멀리 떨어진 데이터센터에 있는 게 아니라, 단말 장치 혹은 게이트웨이와 같은 엑세스 포인트에 가까운 엣지 사이드에 컴퓨팅 능력 및 데이터 처리 능력을 부가함으로써 저지연/초고속컴퓨팅의 실현을 가능케 한다. 이러한 엣지 컴퓨팅의 종류로는 Mobile edge computing, Fog computing, Cloudlet computing이 있으며, 본 논문에서는 엣지 컴퓨팅을 실제 구현/구축하기 위해 현존하는 오픈 소스 플랫폼들에 대해 초점을 맞추고 분석한다. 각 오픈 소스 엣지 플랫폼에 대해 구조 및 특징들을 체계적으로 묘사하고 비교 분석함으로써 오픈 소스 엣지 플랫폼을 이용하여 실제 엣지 노드를 구축 하고자 하는 산업계 엔지니어들에게 사용 사례에 부합한 최선의 엣지 플랫폼을 선택 할 수 있도록 하나의 제반 지식을 제공하고자 한다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1997년도 한국자동제어학술회의논문집; 한국전력공사 서울연수원; 17-18 Oct. 1997
• /
• pp.480-483
• /
• 1997

A fossil power plant can be modeled by a lot of algebraic equations and differential equations. When we simulate a large, complicated fossil power plant by a computer such as workstation or PC, it takes much time until overall equations are completely calculated. Therefore, new processing systems which have high computing speed is ultimately needed to develope real-time simulators. Vital points of real-time simulators are accuracy, computing speed, and deadline observing. In this paper, we present a enhanced strategy in which we can provide powerful computing power by parallel processing of DSP processors with communication links. We designed general purpose DSP modules, and a VME interface module. Because the DSP module is designed for general purpose, we can easily expand the parallel system by just connecting new DSP modules to the system. Additionally we propose methods about downloading programs, initial data to each DSP module via VME bus, DPRAM and processing sequences about computing and updating values between DSP modules and CPU30 board when the simulator is working.

• 한국CDE학회논문집
• /
• 제17권5호
• /
• pp.342-352
• /
• 2012

As a futuristic construction model, building information model (BIM) based project management system (PMIS) and mobile BIM simulator apps have been showing visible sign. However, researches on the BIM based 3D simulator using mobile device are hard to find result from limitation of mobile device (slow speed at huge 3D file, display size, and etc.) and undefined standard of business processes. Therefore, this research aims at studying application of mobile BIM apps based on cloud computing. Total 8 BIM cloud apps were selected and analyzed in the 5 application feasibility characteristics (speed, view, inquiry, markup, and usability). This research would be essential phase to construct BIM based mobile project management system using cloud computing in the future.

• Journal of Electrical Engineering and Technology
• /
• 제3권4호
• /
• pp.552-558
• /
• 2008

Optimal design of the direct-driven Permanent Magnet(PM) wind generator, combined with F.E.A(Finite Element Analysis) and Genetic Algorithm(GA), has been performed to maximize the Annual Energy Production(AEP) over the entire wind speed characterized by the statistical model of wind speed distribution. Particularly, the proposed parallel computing via internet web service has contributed to reducing excessive computing times for optimization.

• 전자공학회논문지A
• /
• 제32A권12호
• /
• pp.184-197
• /
• 1995

In this paper, high speed IC package-to-package interconnections are modeled as lossless multiconductor transmission lines operating in the TEM mode. And, three mathematical algorithms for computing electrical parameters of the lossless multiconductor transmission lines are described. A semi-analytic Green's function method is used in computing per unit length capacitance and inductance matrices, a matrix square root algorithm based on the QR algorithm is used in computing a characteristic impedance matrix, and a matrix algorithm based on the theory of M-matrix is used in computing a diagonally matched load impedance matrix. These algorithms are implemented in a computer program DIME (DIagonally Matched Load Impedance Extractor) which computes electrical parameters of the lossless multiconductor transmission lines. Also, to illustrate the concept of design database for high speed IC package-to-package interconnection, a database for the multi conductor strip transmission lines system is constructed. This database is constructed with a sufficiently small number of nodes using the multi-dimensional cubic spline interpolation algorithm. The maximum interpolation error for diagonally matched load impedance matrix extraction from the database is 1.3 %.

• 한국컴퓨터그래픽스학회논문지
• /
• 제6권3호
• /
• pp.35-40
• /
• 2000

There are two problems in H.26X compression technique. One is compressing time in encoding process and the other is degradation of the decoded video quality due to high compression rate. For transferring moving pictures in real-time, it is required to adopt massively high compression. In this case, there are a lot of losses of an original video data and that results in degradation of quality. Especially degradation called by blocking artifact may be produced. The blocking artifact effect is produced by DCT-based coding techniques because they operate without considering correlation between pixels in block boundaries. So it represents discontinuity between adjacent blocks. This paper describes methods of quality compensation for H.26x decoded data and enhancing encoding speed for real-time operation. Our goal of the quality compensation is not to make the decoded video identical to a original video but to make it perceived better through human eyes. We suggest an algorithm that reduces block artifact and clears decoded video in decoder. To enhance encoding speed, we adopt new four-step search algorithm. As shown in the experimental result, the quality compensation provides better video quality because of reducing blocking artifact. And then new four-step search algorithm with $MMX^{TM}$ implementation improves encoding speed from 2.5 fps to 17 fps.

• Journal of Computing Science and Engineering
• /
• 제1권2호
• /
• pp.195-210
• /
• 2007

Dwarf is a highly compressed structure, which compresses the cube by eliminating the semantic redundancies while computing a data cube. Although it has high compression ratio, Dwarf is slower in querying and more difficult in updating due to its structure characteristics. We all know that the original intention of data cube is to speed up the query performance, so we propose two novel clustering methods for query optimization: the recursion clustering method which clusters the nodes in a recursive manner to speed up point queries and the hierarchical clustering method which clusters the nodes of the same dimension to speed up range queries. To facilitate the implementation, we design a partition strategy and a logical clustering mechanism. Experimental results show our methods can effectively improve the query performance on data cubes, and the recursion clustering method is suitable for both point queries and range queries.

• 한국생물물리학회:학술대회논문집
• /
• 한국생물물리학회 2003년도 정기총회 및 학술발표회
• /
• pp.65-65
• /
• 2003

After many genome projects, algorithms and software to process explosively growing biological information have been developed. To process huge amount of biological information, high performance computing equipments are essential. If we use the remote resources such as computing power, storages etc., through a Grid to share the resources in the Internet environment, we will be able to obtain great efficiency to process data at a low cost. Here we present the performance improvement of the protein secondary structure prediction (PSIPred) by using the Grid platform, distributing protein sequence data on the Grid where each computer node analyzes its own part of protein sequence data to speed up the structure prediction. On the Grid, genome scale secondary structure prediction for Mycoplasma genitalium, Escherichia coli, Helicobacter pylori, Saccharomyces cerevisiae and Caenorhabditis slogans were performed and analyzed by a statistical way to show the protein structural deviation and comparison between the genomes. Experimental results show that the Grid is a viable platform to speed up the protein structure prediction and from the predicted structures.