• 기술혁신연구
• /
• 제23권2호
• /
• pp.33-54
• /
• 2015

국가차원의 슈퍼컴퓨팅 성능수요 예측은 슈퍼컴퓨터를 활용하는 계산과학분야의 연구자나 연구개발 인프라를 구축 운영하고 있는 전문기관, 과학기술 인프라구축을 주도할 정부기관에 있어서 매우 중요한 정보이다. 본 연구는 그동안 진행되었던 슈퍼컴퓨터 성능관련 예측활동 분석을 통해 과학기술 역량에 영향을 미치는 요인들을 도출하고 이를 슈퍼컴퓨터 기술진보 추세에 적용한 복합 예측모형을 제안하였다. 횡단면분석에서는 슈퍼컴퓨팅 성능에 영향을 미칠 것으로 판단되는 GDP, GERD, 연구원수, SCI논문수를 고려한 다중회귀분석을 수행하였다. 그리고 횡단면분석 결과에 Top500 자료의 성능(Rmax)값을 이용한 시계열분석을 통해 도출된 기간별 기술진보율을 곱하여 슈퍼컴퓨터의 성능을 예측하였다. 제안된 예측모형을 바탕으로 세계 슈퍼컴퓨터 500위의 시계열자료를 이용하여 한국이 2016년에 보유해야 할 슈퍼컴퓨터 성능규모를 예측하였다. 횡단면분석과 기술진보율을 적용하여 2016년 한국의 슈퍼컴퓨팅 성능수요를 예측해본 결과 현재의 추세를 이용할 경우 15~30PF 정도, 목표 국가수준의 추세를 이용할 때 20~40PF 정도의 컴퓨팅 역량이 필요할 것으로 예측되었다. 이 결과는 단순 회귀분석을 적용한 결과인 9.6PF와 횡단면분석을 적용한 결과인 2.5PF와 큰 차이를 나타내었다.

• Journal of applied mathematics & informatics
• /
• 제3권2호
• /
• pp.217-234
• /
• 1996

This report introduces a faster parallel LU decomposi-tion algorithm that gives a speedup almost equal to the number of nodes used. The new algorithm takes an advantage of an important C feature that lays out a matrix using a row major scheme and is based on the currently widely used LU decomposition algorithm with one major modification to eliminate most of the communication overhead. Empirical results are included in this report. For example solving a dense matrix that contains 100,000,000 elements gives a speedup of 50 when executed on 50 nodes of an intel Paragon in parallel.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 컴퓨터소사이어티 추계학술대회논문집
• /
• pp.45-48
• /
• 2003

We can easily buy network system for high performance micro-processor, progress computer architecture is caused of high bandwidth and low delay time. Coupling PC-based commodity technology with distributed computing methodologies provides an important advance in the development of single-user dedicated systems. Lately Network is joined PC or workstation by computers of high performance and low cost. Than it make intensive that Cluster system is resembled supercomputer. Unix, Linux, BSD, NT(Windows series) can use Cluster system OS(operating system). I'm chosen linux gain low cost, high performance and open technical documentation. This paper is benchmark performance of Beowulf clustering by UltraSPARC-1K(64bit-RISC processor). Benchmark tools use MPI(Message Passing Interface) and NetPIPE. Beowulf is a class of experimental parallel workstations developed to evaluate and characterize the design space of this new operating point in price-performance.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 1998년도 춘계학술발표 논문집
• /
• pp.287-292
• /
• 1998

The purpose of the subjects described in this paper is to share the inforamation and utilize expensive equipmints and supercomputer at remote site ghrough the latist telecommunication and control technology without tempoual and spatial limitation. The major subjects areabstracted as two. One for developing computer supported cooperative works and experimintal systims which make it possible that researchers discuss about the result of collaborated research and exchange data. And the other for connecting the university, research institutes and industries in the country to information Highway applying developed systim. The gargeted filds of this project are Machinery Design, Chemistry and Chemical Engineering and wil be expanded to polymer Engineering, Medicine, Agricultural Chemicals.

• Advances in Computational Design
• /
• 제4권4호
• /
• pp.343-365
• /
• 2019

Among the themes related to earthquake countermeasures at construction sites, those for tower cranes are particularly important. An accident involving the collapse of a crane during the construction of a skyscraper has serious consequences, such as human injury or death, enormous repair costs, and significant delays in construction. One of the causes of deadly tower crane collapses is the destruction of the site joints of the tower crane mast. This paper proposes a new design method by static elastoplastic finite element analysis using a supercomputer for the design of the end plate-type tensile bolted joints, which are generally applied to the site joints of a tower crane mast. This new design method not only enables highly accurate and reliable joint design but also allows for a design that considers construction conditions, such as the introduction of a pre-tension axial force on the bolts. By applying this new design method, the earthquake resistance of tower cranes will undoubtedly be improved.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.118-121
• /
• 2021

슈퍼컴퓨터는 수백~수천 노드의 컴퓨팅 자원이 연결되어 복잡한 계산이나 대규모 데이터를 병렬 처리하며 일부 노드에서 발생하는 예상치 못한 문제는 전체적인 시스템 성능 저하로 이어질 수 있기 때문에 슈퍼컴퓨터 모니터링과 성능 분석은 슈퍼컴퓨터를 구축하고 운영하는데 필수적인 요소로 볼 수 있다. 본 논문에서는 오픈소스 소프트웨어인 Performance Co-Pilot, Bpftrace, Grafana를 활용한 슈퍼컴퓨터 모니터링 및 성능분석 시스템 구축 방안을 제안하였으며 이를 통해서 확장가능하면서도 유연한 구조의 슈퍼컴퓨터 모니터링 및 성능 분석이 가능함을 보였다.

• 인터넷정보학회논문지
• /
• 제17권3호
• /
• pp.129-136
• /
• 2016

Crucial aspects to successfully realizing the re-emergence of a contemporary and sustainable supercomputing community in South Korea will involve the devoted efforts and support from key government and R&D organizations. We suggest various supplementation plans regarding the roles of support for the statutory plan. This includes the committee and the plans which are often missing necessary support systems that help competent ministries to plan properly according to the missions of the research center. This dissertation suggests that adjustment in the HPC trends will depend upon exposing and correcting problems in the law as well as overall improvement of the law. Also, the total development of a super computing market is necessary. The results of these guidelines will create a spread of demand for supercomputing for national IT resource sharing, and will foster the development of supercomputer specialists worldwide. Other major end results include significant increases in research productivity and increased rates of product development.

• Genomics & Informatics
• /
• 제20권1호
• /
• pp.14.1-14.4
• /
• 2022

Permutation testing is a robust and popular approach for significance testing in genomic research that has the advantage of reducing inflated type 1 error rates; however, its computational cost is notorious in genome-wide association studies (GWAS). Here, we developed a supercomputing-aided approach to accelerate the permutation testing for GWAS, based on the message-passing interface (MPI) on parallel computing architecture. Our application, called MPI-GWAS, conducts MPI-based permutation testing using a parallel computing approach with our supercomputing system, Nurion (8,305 compute nodes, and 563,740 central processing units [CPUs]). For 10⁷ permutations of one locus in MPI-GWAS, it was calculated in 600 s using 2,720 CPU cores. For 10⁷ permutations of ~30,000-50,000 loci in over 7,000 subjects, the total elapsed time was ~4 days in the Nurion supercomputer. Thus, MPI-GWAS enables us to feasibly compute the permutation-based GWAS within a reason-able time by harnessing the power of parallel computing resources.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2016년도 춘계학술대회
• /
• pp.554-557
• /
• 2016

최근, 슈퍼컴퓨터 시스템은 교육, 의료, 국방 등은 물론 계산과학 시뮬레이션까지 다양한 분야에서 활용되고 있다. 특히, 슈퍼컴퓨터, 클라우드 등 이종의 계산 자원을 연동하는 시뮬레이션 환경은 시간, 공간적 한계를 극복하는 차세대 연구 환경으로서 다양한 응용 분야에서 큰 각광을 받고 있다. 그러나 기존의 이종 슈퍼컴퓨팅 환경은 API(Application Programming Interface)의 부재로 터미널 접근을 통해 계산 작업을 수행하는 단조로운 형태를 띠고 있으며, 이는 이종 컴퓨팅 자원의 접근성 및 활용성을 저해하는 주요 요소이다. 이러한 문제점을 해결하고, 슈퍼컴퓨팅 서비스의 다양성을 확보하기 위해, 본 논문에서는 계산과학 시뮬레이션을 위한 웹 기반 HPC (High Performance Computing) 작업 관리 프레임워크의 설계 및 구현에 대해 기술한다. 제안한 프레임워크는 슈퍼컴퓨터, 클라우드, 그리드 등의 다양한 이기종 자원을 활용하여 시뮬레이션 계산 작업을 수행할 수 있도록 플러그-인 기반 표준 인터페이스 및 번들 플러그-인을 제공하며, 플러그-인 관리자를 통해 손쉽게 자원 확장이 가능하다. 또한 시뮬레이션 작업의 생성, 제출, 제어, 상태 모니터링 등의 전주기적인 작업 라이프사이클 관리를 위해 HTTP 표준 방식의 RESTful endpoints들을 제공한다.

• Journal of Information Processing Systems
• /
• 제2권2호
• /
• pp.101-106
• /
• 2006

Grid computing is an emerging technology that enables global resource sharing. In Korea, the $K^*$Grid provides an extremely powerful research environment to both industries and academia. As part of the $K^*$Grid project, we have constructed, together with the Korea Institute of Science and Technology Information and a number of domestic universities, a supercomputer Grid test bed which connects several types of supercomputers based on the globus toolkit. To achieve efficient networking in this Grid testbed, we propose a novel method of available bandwidth measurement, called Decoupled Capacity measurement with Initial Gap (DCIG), using packet trains. DCIG can improve the network efficiency by selecting the best path among several candidates. Simulation results show that DCIG outperforms previous work in terms of accuracy and the required measurement time. We also define a new XML schema for DCIG request/response based on the schema defined by the Global Grid Forum (GGF) Network Measurement Working Group (NM-WG).