A compilation tool for automated mapping of algorithms onto FPGA custom computing machines
Başlık çevirisi mevcut değil.
- Tez No: 400117
- Danışmanlar: DR. CLAY S. GLOSTER, DR. WİNSER E. ALEXANDER
- Tez Türü: Doktora
- Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2002
- Dil: İngilizce
- Üniversite: North Carolina State University
- Enstitü: Yurtdışı Enstitü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 202
Özet
Özet yok.
Özet (Çeviri)
A Compilation Tool for Automated Mapping of Algorithmsonto FPGA-Based Custom Computing Machines. (Under the directions of Dr.Clay S. Gloster and Dr. Winser E. Alexander).Adaptive computing, also known as Recongurable Computing (RC), is aeld that combines hardware and software data processing platforms. RCsystems combine the flexibility of General Purpose Processors (GPP) withthe speed of application specic processors [1, 2]. In a typical recongurablecomputer, computationally intensive portions of algorithms are executed on FieldProgrammable Gate Arrays (FPGA) for enhanced performance.Although RC systems oer signicant performance advantages over GPPs,they have a few disadvantages. RC systems require more application developmenttime than GPPs. Also, RC system designers need to be knowledgeable in theareas of hardware and software system design. Since each application is dierentin terms of data inputs, outputs, and the method of processing data, designersare required to design a specic RC implementation for each specic problem.Our major contribution in this research is the development of adesign automation tool called the Recongurable Computing Com-pilation Tool (RCCT) to address the problems mentioned above. Inaddition, this tool was designed to automate the process of mapping applicationsonto RC systems, and to provide the potential performance benets of RCsystems to typical software programmers. The nal version of the tool containsfour components: The RC Compiler, the Module Library, the Loader and theSimulator. Our contributions also includes a novel assembly languageinstruction set for the modules and a session le format (a newassembly language program format for RC systems).The tool was tested on several applications to demonstrate its eectiveness.Among the selected applications were matrix multiplication, and some imageprocessing algorithms such as 3-D Image correlation. We compared the executiontimes of the applications when they were run on dierent GPPs to dierent RCcongurations to demonstrate the tool's eectiveness.Our results showed that the tool is able to enhance the performance of theapplications by mapping portions of them to the RC systems. Simulationswith the tool showed that when the user applications are mapped to the RCsystems, signicant speedups (around 10 times to 100 times) can be attainedfor the mapped sections of the applications. We also noticed that the designand implementation time of the RC versions of the applications were reducedsignicantly. With the tool, the RC versions of the applications were developed,in a matter of a few hours. No special skills are needed to map applications to theRC systems using RCCT if the required hardware modules are readily available.
Benzer Tezler
- Developing algorithm for automatic detection of caves using unmanned aerial vehicle data
İnsansız hava araçlarından elde edilen verilerle otomatik mağara tespiti yapabilen algoritmanın geliştirilmesi
MUSTAFA BÜNYAMİN SAĞMAN
Yüksek Lisans
İngilizce
2023
Savunma ve Savunma Teknolojileriİstanbul Teknik ÜniversitesiSavunma Teknolojileri Ana Bilim Dalı
DOÇ. DR. ORKAN ÖZCAN
- Exploring fusion models in computer vision for medical image computing
Başlık çevirisi yok
DUYGU SARIKAYA
Doktora
İngilizce
2017
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve KontrolState University of New York at BuffaloDr. JASON J. CORSO
- Zemin sıvılaşmasının hava fotoğraflarının değerlendirilmesi yöntemi ile tespiti
The Determination of liguefaction with the compilation method of acrial photographs
VEYSEL OKAN ATAK
- Doğrusal ve doğrusal olmayan metotlarla bir adım ileri rüzgar şiddeti öngörüsü
One step ahead wind speed prediction with linear and nonlinear methods
İREM DAMLA NACAR
Yüksek Lisans
Türkçe
2019
Enerjiİstanbul Teknik ÜniversitesiEnerji Bilim ve Teknoloji Ana Bilim Dalı
DR. ÖĞR. ÜYESİ BURAK BARUTÇU