• 제목/요약/키워드: Data Locality

검색결과 237건 처리시간 0.023초

장소인식멀티센서스마트 환경을위한 데이터 퓨전 모델 (Locality Aware Multi-Sensor Data Fusion Model for Smart Environments)

  • 와카스 나와즈;무하머디 파힘;이승룡;이영구
    • 한국정보처리학회:학술대회논문집
    • /
    • 한국정보처리학회 2011년도 춘계학술발표대회
    • /
    • pp.78-80
    • /
    • 2011
  • In the area of data fusion, dealing with heterogeneous data sources, numerous models have been proposed in last three decades to facilitate different application domains i.e. Department of Defense (DoD), monitoring of complex machinery, medical diagnosis and smart buildings. All of these models shared the theme of multiple levels processing to get more reliable and accurate information. In this paper, we consider five most widely acceptable fusion models (Intelligence Cycle, Joint Directors of Laboratories, Boyd control, Waterfall, Omnibus) applied to different areas for data fusion. When they are exposed to a real scenario, where large dataset from heterogeneous sources is utilize for object monitoring, then it may leads us to non-efficient and unreliable information for decision making. The proposed variation works better in terms of time and accuracy due to prior data diminution.

Enhancing GPU Performance by Efficient Hardware-Based and Hybrid L1 Data Cache Bypassing

  • Huangfu, Yijie;Zhang, Wei
    • Journal of Computing Science and Engineering
    • /
    • 제11권2호
    • /
    • pp.69-77
    • /
    • 2017
  • Recent GPUs have adopted cache memory to benefit general-purpose GPU (GPGPU) programs. However, unlike CPU programs, GPGPU programs typically have considerably less temporal/spatial locality. Moreover, the L1 data cache is used by many threads that access a data size typically considerably larger than the L1 cache, making it critical to bypass L1 data cache intelligently to enhance GPU cache performance. In this paper, we examine GPU cache access behavior and propose a simple hardware-based GPU cache bypassing method that can be applied to GPU applications without recompiling programs. Moreover, we introduce a hybrid method that integrates static profiling information and hardware-based bypassing to further enhance performance. Our experimental results reveal that hardware-based cache bypassing can boost performance for most benchmarks, and the hybrid method can achieve performance comparable to state-of-the-art compiler-based bypassing with considerably less profiling cost.

직접 사상 캐쉬의 캐쉬 실패율을 감소시키기 위한 성김도 정책 (Determination of a Grain Size for Reducing Cache Miss Rate of Direct-Mapped Caches)

  • 정인범;공기석;이준원
    • 한국정보과학회논문지:시스템및이론
    • /
    • 제27권7호
    • /
    • pp.665-674
    • /
    • 2000
  • 높은 캐쉬 구역성을 나타내는 데이터 병렬 프로그램들에서 그레인 크기의 선정은 캐쉬 성능에 커다란 영향을 마친다. 선택된 그레인 크기가 프로세서들 사이에 균등한 부하 배분을 제공하더라도 내재하는 캐쉬 효과를 무시한 그레인 크기는 하나의 프로세서에 할당된 그레인들 사이에 주소 간섭을 발생되게 한다. 이런 주소 간섭은 캐쉬 충돌 실패를 발생하기 때문에 캐쉬 구역성에 부정적 영향을 미치게 된다. 이러한 문제점을 해결하기 위해서 본 논문에서는 직접 사상 캐쉬의 특정을 바탕으로 캐쉬 크기와 프로세서 개수들로부터 최적 그레인 크기를 유도한다. 제안된 방법은 캐쉬 공간 내에서 그레인들이 동일한 주소로 사상 되지 않게 하므로 캐쉬 충돌 실패를 감소시킨다. 모의 시험 결과는 제안된 최적 그레인 크기는 직접 사상 캐쉬상에서 캐쉬 실패를 줄이므로 시험된 데이터 병렬 프로그램들의 성능을 개선시킴을 보인다.

  • PDF

분산환경에서 대용량 공간데이타의 선인출 전략 (A Prefetch Policy for Large Spatial Data in Distributed Environment)

  • 박동주;이석호;김형주
    • 한국정보과학회논문지:소프트웨어및응용
    • /
    • 제26권12호
    • /
    • pp.1404-1417
    • /
    • 1999
  • 웹과 같은 분산 환경에서는, 웹 브라우저 상에서 SQL 형식의 공간 질의를 수행시키는 것과 또한 서버로부터 그 질의 결과를 보는 것이 가능하다. 그러나, 격자 이미지(raster image)와 같은 대용량 공간 데이타를 포함하는 질의 결과를 웹 브라우징할 때 발생하는 많은 문제점들 중에서, 사용자 응답 시간의 지연은 매우 중요한 문제이다. 본 논문에서는 사용자의 재요청(callback) 접근 패턴이 공간적 근접성(spatial locality)을 따른다는 가정하에서의, 사용자 응답 시간을 최소화하기 위한 새로운 프리페치(prefetch) 전략에 대해서 서술한다. 본 논문의 프리페치 전략은 다음과 같이 요약될 수 있다. 첫째, 프리페치 알고리즘은 사용자의 접근 패턴을 잘 반영하는 힐버트 곡선(Hilbert-curve) 모델을 바탕으로 한다. 둘째, 프리페치 전송 비용을 줄이기 위해서 사용자의 재요청 시간 간격(think time)을 이용한다. 본 논문에서는, 힐버트 곡선을 이용한 프리페치 전략의 성능 평가를 위해서 다양한 실험을 하였으며, 그 결과로 프리페치를 하지 않는 방식보다 높은 성능 향상이 있음을 보인다.Abstract In distributed environment(e.g., WWW), it would be possible for the users to submit SQL-like spatial queries and to see their query results from the server on the Web browser. However, of many obstacles which result from browsing query results including large spatial data such as raster image, the delay of user response time is very critical. In this paper we present a new prefetch policy which can alleviate user response time on the assumption that user's callback access pattern has spatial locality. Our prefetch policy can be summerized as follows: 1) our prefetch algorithm is based on the Hibert-curve model which well replects user's access pattern, and 2) it utilizes user's callback interval to reduce prefetch network transmission cost. In this paper we conducted diverse experiments to show that our prefetch policy achieves higher performance improvement compared to other non-prefetch methods.

P2P 스트리밍 시스템을 위한 TTL 기반 피어 그룹핑 기법 (A TTL-based Peer Grouping Scheme for P2P Streaming Systems)

  • 김태영;김은삼
    • 한국컴퓨터정보학회논문지
    • /
    • 제17권1호
    • /
    • pp.151-159
    • /
    • 2012
  • 본 논문에서는 P2P 기반의 IPTV 시스템에서 TTL을 이용해서 지역성을 반영하는 피어 그룹핑 기법을 제안한다. 이 기법은 두 피어 사이의 경유 라우터 수를 최소화하도록 이웃 피어를 선정하여 중간 라우터들이 전송해야 하는 데이터 양이 감소한 만큼 네트워크 대역폭을 절약할 수 있다. 또한 본 논문에서 제안하는 TTL 기반의 피어 그룹핑 기법이 기존의 RTT 기반의 기법에 비해 성능이 향상되는 것을 보이기 위해 다양한 시뮬레이션 시험을 수행하였다. 실험 결과는 본 논문에서 제안하는 TTL 기반의 피어 그룹핑 기법이 전체 라우터들이 전송하는 데이터 및 제어패킷 용량뿐만 아니라 지터 비율도 감소시킬 수 있다는 것을 보인다.

Web access prediction based on parallel deep learning

  • Togtokh, Gantur;Kim, Kyung-Chang
    • 한국컴퓨터정보학회논문지
    • /
    • 제24권11호
    • /
    • pp.51-59
    • /
    • 2019
  • 웹에서 정보 접근에 대한 폭발적인 주문으로 웹 사용자의 다음 접근 페이지를 예측하는 필요성이 대두되었다. 웹 접근 예측을 위해 마코브(markov) 모델, 딥 신경망, 벡터 머신, 퍼지 추론 모델 등 많은 모델이 제안되었다. 신경망 모델에 기반한 딥러닝 기법에서 대규모 웹 사용 데이터에 대한 학습 시간이 엄청 길어진다. 이 문제를 해결하기 위하여 딥 신경망 모델에서는 학습을 여러 컴퓨터에 동시에, 즉 병렬로 학습시킨다. 본 논문에서는 먼저 스파크 클러스터에서 다층 Perceptron 모델을 학습 시킬 때 중요한 데이터 분할, shuffling, 압축, locality와 관련된 기본 파라미터들이 얼마만큼 영향을 미치는지 살펴보았다. 그 다음 웹 접근 예측을 위해 다층 Perceptron 모델을 학습 시킬 때 성능을 높이기 위하여 이들 스파크 파라미터들을 튜닝 하였다. 실험을 통하여 논문에서 제안한 스파크 파라미터 튜닝을 통한 웹 접근 예측 모델이 파라미터 튜닝을 하지 않았을 경우와 비교하여 웹 접근 예측에 대한 정확성과 성능 향상의 효과를 보였다.

MapReduce 작업처리시간 단축을 위한 선 정렬 기반 태스크 스케줄링 기법 (Pre-arrangement Based Task Scheduling Scheme for Reducing MapReduce Job Processing Time)

  • 박정효;김준상;김창현;이원주;전창호
    • 한국컴퓨터정보학회논문지
    • /
    • 제18권11호
    • /
    • pp.23-30
    • /
    • 2013
  • 본 논문에서는 MapReduce 작업처리시간을 줄일 수 있는 선 정렬 기반 태스크 스케줄링 기법을 제안한다. 태스크와 그 태스크가 처리할 데이터가 동일 노드에 존재하지 않으면 해당 태스크는 다른 노드로부터 데이터를 전송받아 처리한다. 이때 전송시간으로 인해 MapReduce의 작업처리시간이 증가하는 문제점이 발생한다. 이러한 문제점을 해결하기 위해 본 논문에서는 두 단계로 태스크를 스케줄링한다. 첫 번째 단계에서는 데이터 지역성이 높은 순으로 태스크를 노드 리스트에 정렬한다. 두 번째 단계에서는 데이터의 위치정보를 이용하여 태스크들이 데이터 지역성을 높일 수 있도록 교환하여 스케줄링한다. 본 논문에서는 제안한 스케줄링 기법의 성능평가를 위해 소규모 Hadoop 클러스터를 구현하여 실험하였다. 제안한 기법을 적용하였을 때 작업처리시간이 약 18% 감소하였으며 데이터가 저장된 노드에 할당되지 않은 태스크 수는 약 25% 감소하였다.

메모리 지연을 감추는 기법들 (Memory Latency Hiding Techniques)

  • 기안도
    • 전자통신동향분석
    • /
    • 제13권3호통권51호
    • /
    • pp.61-70
    • /
    • 1998
  • The obvious way to make a computer system more powerful is to make the processor as fast as possible. Furthermore, adopting a large number of such fast processors would be the next step. This multiprocessor system could be useful only if it distributes workload uniformly and if its processors are fully utilized. To achieve a higher processor utilization, memory access latency must be reduced as much as possible and even more the remaining latency must be hidden. The actual latency can be reduced by using fast logic and the effective latency can be reduced by using cache. This article discusses what the memory latency problem is, how serious it is by presenting analytical and simulation results, and existing techniques for coping with it; such as write-buffer, relaxed consistency model, multi-threading, data locality optimization, data forwarding, and data prefetching.

주문형 전자신문 시스템에서 사용자 접근패턴을 이용한 기사 프리패칭 기법 (Article Data Prefetching Policy using User Access Patterns in News-On-demand System)

  • 김영주;최태욱
    • 한국정보처리학회논문지
    • /
    • 제6권5호
    • /
    • pp.1189-1202
    • /
    • 1999
  • As compared with VOD data, NOD article data has the following characteristics: it is created at any time, has a short life cycle, is selected as not one article but several articles by a user, and has high access locality in time. Because of these intrinsic features, user access patterns of NOD article data are different from those of VOD. Thus, building NOD system using the existing techniques of VOD system leads to poor performance. In this paper, we analysis the log file of a currently running electronic newspaper, show that the popularity distribution of NOD articles is different from Zipf distribution of VOD data, and suggest a new popularity model of NOD article data MS-Zipf(Multi-Selection Zipf) distribution and its approximate solution. Also we present a life cycle model of NOD article data, which shows changes of popularity over time. Using this life cycle model, we develop LLBF (Largest Life-cycle Based Frequency) prefetching algorithm and analysis he performance by simulation. The developed LLBF algorithm supports the similar level in hit-ratio to the other prefetching algorithms such as LRU(Least Recently Used) etc, while decreasing the number of data replacement in article prefetching and reducing the overhead of the prefetching in system performance. Using the accurate user access patterns of NOD article data, we could analysis correctly the performance of NOD server system and develop the efficient policies in the implementation of NOD server system.

  • PDF

VIDEO GAME CULTURE AND INTERACTIVITY -An exploration of digital interactive media through a metaphorical approach to video game culture-

  • 우탁
    • 한국게임학회지
    • /
    • 제6권1호
    • /
    • pp.70-72
    • /
    • 2009
  • This research is focused on defining interaction within the context of digital media and creating a multi cultural definition of interactivity. The concept of multi digital culture and a definition of interaction in digital media have often been overlooked by other researchers and this has caused the emergence of many different notions on this issue. As a result of these varied notions of the concept, public confusion has arisen regarding interactivity. The main purpose of this research is to find a suitable multi definition of interaction through examining local digital culture. In order to analogise multi digital culture, the video game culture is employed as a metaphor to interpret local digital culture. The reason for this is that a specific national culture can be easily identified within the video game culture. Four countries, South Korea, Japan, the U.S. and the UK have been chosen for comparison purposes. Case study, questionnaire and publicly accessible video game related data, such as, video game charts, are used for formalising and analysing unique local digital culture. The Heyri POP UP IMAGE Festival, S. Korea, was also used as a pilot study, with some of the above research methods being employed to analyse South Korean digital culture. In relation to western cases, interview and questionnaire were primarily used. The data from the case countries was carefully compared and analysed and then it became the basis of a theory of multi definition of interaction in digital media. The case study employed the cultural metaphor for this research and in addition video game culture related questionnaires and interviews with experts of interactive art genre, regarding new notions of digital interaction were utilised. The survey was conducted simultaneously in the four different cultural case nations of this research. Twenty respondents from each case nation participated in the survey, in order to investigate firstly, the existence of 'local digital culture' and secondly, the trends and phenomena of 'digital culture' in these four different 'local digital cultural areas'. In terms of interviews with experts of the interactive art genre, these were focused on obtaining their understanding of contemporary digital culture in their research. Using gathered data from the observation of local digital culture, the basic theory of interaction and the terminology of interaction are reformed. Localised definitions of interaction on digital media, control based interaction and communication based interaction are presented, in order to identify a 'locality' in terms of various contemporary digital cultures. As a result of analysing digital culture, new definitions of 'multi definition of digital interaction' were formulated. As mentioned above, 'control' and 'communication' based interaction were initiated, based on 'user to media' relationships. Based on the degree of physical interaction, 'liminal' and 'transitive' interactions were initiated. Less physical digital interaction is named 'liminal' interaction and more physical digital interaction is named' transitive' interaction. These new definitions of interaction were applied to the real world examples of uses of digital interaction, such as, digital interactive installation artworks and video games. The newly defined meaning of digital interaction can be applied to analysing digital interactive installation artworks and possibly indicate their future development and the prospects of future electronic games. Three leading digital interactive artists were selected for this analysis and their works were studied in terms of the implementation of 'multi definition of digital interaction'. Throughout these processes, the meaning of 'communication' in digital interactive media was emphasised. Many of the selected artists' digital installations were focused on 'communication' or 'interaction between each user through digital media', rather than the concept of 'control' in digital interaction, otherwise termed, 'communication with digital media'. In their artworks, interaction between each audience was digitally engaged within the physical interactive environment which was created by the digital media. Both the audience's action and all the reaction throughout the interaction between the audiences, triggered the digital media' s reaction. This audience-audience-media interaction is the key to understanding the concept of 'communication' in physical digital media and it is the main interactive concept upon which the selected digital interactive installation artists for this research and many other artists from similar fields, are concentrating their efforts. In the case of the video game, a similar trend was noticed to that of digital interactive installations. Based on this research's 'multi definition of digital interaction', the video game has evolved from the early stage of being conventional game, which was focused on control based interaction, to the on-line game which was focused on communication based interaction, to physical interactive games, such as, Nintendo Wii, which are focused on more physical interaction and finally, the ubiquitous interactive game, which is mainly concentrated on the concept of 'communication' in physical digital interaction. It is possible that this evolution of the video game concept of interaction is comparable to the progress of digital interactive artworks. This view is based on the fact that both genres show evidence that they are developing in the direction of the concept of 'communication', in terms of physical digital interaction. The important emphasis of this research's results is 'locality' and 'communication' in physical digital interaction. The existence of different digital culture trends, which were assessed by the 'multi definition of digital interaction', can explain the concept of 'locality' in digital interaction. This meaning of 'locality' may assist in understanding contemporary digital culture and can reduce possible misunderstanding as regards 'local' digital culture. In the application of the concept of digital interaction to the field of either artworks or video games, it is possible to form the opinion that an innovative concept of physical digital interaction is 'communication' within this context. This concept and its applications can improve the potential of both digital interactive culture and technology.

  • PDF