Categorization and visualization of parallel programming systems
Paralel programlama sistemlerinin sınıflandırılması ve grafik gösterimi
- Tez No: 166806
- Danışmanlar: PROF. DR. EMRE HARMANCI
- Tez Türü: Yüksek Lisans
- Konular: Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrol, Computer Engineering and Computer Science and Control
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2005
- Dil: İngilizce
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Bilgisayar Mühendisliği Ana Bilim Dalı
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 91
Özet
differences of these styles. Algorithms are implemented in OpenMP and MPI, performance of both programs are measured on the SMP Cluster of Aachen University, Germany and on the Beowulf Cluster of Ulakbim, Ankara. xin
Özet (Çeviri)
CATEGORIZATION AND VISUALIZATION OF PARALLEL PROGRAMMING SYSTEMS ABSTRACT Parallel computing, also called high-performance computing, refers to solving problems faster by using multiple processors simultaneously. Parallel computing became popular in the late 1980s and increased its popularity with the continual desire for more computing power. Nowadays, almost every computationally-intensive problem that one. could imagine, like the simulation of water levels in the rivers, chemical or physical problems, or astronomical simulations is tried to be implemented in parallel. This thesis is aimed at discussing high-performance software for scientific or engineering applications. The term parallel programming systems here means libraries, languages, compiler directives or other means through which a programmer can express a parallel algorithm. To design high performance programs, there are two keys for the programmer: The first is to understand the problem and find a solution for parallelization, and the second is to decide on the right system for the implementation, which requires a good knowledge about existing parallel programming systems. Sometimes, in a parallel application, several hardware/software tools are combined. The programmer, after having understood the problem, has to choose between many systems, some of which are closely related, whereas others have big differences. To give an impression of the variety, a few systems are outlined here, others are explained in the rest of the thesis. This thesis makes four contributions. First it describes and classifies existing parallel programming systems, thus bringing existing surveys up to date. Special emphasis has been given to skeletons and parallel functional programming. Second, it describes a wiki-based web portal for collecting information about most recent systems, which has been developed as part of the thesis. Wiki is a web engine that is fully dynamic and the content can be enriched by the users. Third, it reports on an extension of the wiki technology that has been introduced for the representation of the classifications. A special syntax and a visualization tool has been developed. The syntax in which users can add remarks to web pages, is easy to learn and use. The graph visualization tool uses the remarks to generate visual categorizations and clearly show relations between various systems. This syntax and tool allow users to have their own categorization scheme. Fourth, it compares two major programming styles message passing and shared memory with two different algorithms in order show performance xndifferences of these styles. Algorithms are implemented in OpenMP and MPI, performance of both programs are measured on the SMP Cluster of Aachen University, Germany and on the Beowulf Cluster of Ulakbim, Ankara. xin
Benzer Tezler
- Cephede dıştan kullanılan gölgeleme elemanları ve uygulama yöntemlerinin örnekler üzerinden incelenmesi
External shading devices that are applied to facades and examining shading devices' application methods through application examples
NACİYE SENA ÖZYER
Yüksek Lisans
Türkçe
2017
Mimarlıkİstanbul Teknik ÜniversitesiMimarlık Ana Bilim Dalı
DOÇ. DR. MUSTAFA ERKAN KARAGÜLER
- Using machine learning techniques to enhance teaching and performance prediction of students with autism spectrum disorders
Yapay öğrenme yöntemleri ile otizm spektrum bozukluğu olan öğrencilerin öğretiminin ve öğretim performansı tahmininin iyileştirilmesi
AKRAM M.M. RADWAN
Doktora
İngilizce
2016
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrolİstanbul Teknik ÜniversitesiBilgisayar Mühendisliği Ana Bilim Dalı
PROF. DR. ZEHRA ÇATALTEPE
- Cumhurı̇yet Dönemı̇ basınında (1928 - 2000) harı̇ta ı̇çerı̇klerı̇nı̇n kartografı̇k yönden araştırılması, arşı̇vlenmesı̇, sınıflandırılması
Archiving, classification and research on map contents according to cartographic approach in press of the period (1928-2000) of Turkish Republic
PINAR SARIN
Doktora
Türkçe
2021
Jeodezi ve Fotogrametriİstanbul Teknik ÜniversitesiGeomatik Mühendisliği Ana Bilim Dalı
PROF. DR. NESİBE NECLA ULUĞTEKİN
- Türkçe tümcelerin yüklem odaklı anlam ve dilbilgisi çözümlemesi
Grammatical and semantic analysis of turkish sentence based on predicate
İLKNUR DÖNMEZ
Doktora
Türkçe
2016
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrolİstanbul Teknik ÜniversitesiBilgisayar Mühendisliği Ana Bilim Dalı
PROF. DR. EŞREF ADALI
- Level of detail optimization for 3D city models
3 boyutlu kent modelleri için LOD optimizasyonu
BARKIN EMİRHAN GÜLER
Yüksek Lisans
İngilizce
2023
Şehircilik ve Bölge PlanlamaDokuz Eylül ÜniversitesiJeoloji Mühendisliği Ana Bilim Dalı
PROF. DR. MUHAMMED AYDOĞAN