Spelling suggestions: "subject:"eomputer arithmetic"" "subject:"aomputer arithmetic""
51 |
A VLSI algorithm for computing the Euclidean norm of a 3D vector高木, 直史, Takagi, Naofumi 10 1900 (has links)
No description available.
|
52 |
A hardware algorithm for modular multiplication/division高木, 直史, Takagi, Naofumi 01 1900 (has links)
No description available.
|
53 |
Techniques for FPGA neural modelingWeinstein, Randall Kenneth 21 November 2006 (has links)
Neural simulations and general dynamical system modeling consistently push the limits of available computational horsepower. This is occurring for a number of reasons: 1) models are progressing in complexity as our biological understanding increases, 2) high-level analysis tools including parameter searches and sensitivity analyses are becoming more prevalent, and 3) computational models are increasingly utilized alongside with biological preparations in a dynamic clamp configuration. General-purpose computers, as the primary target for modeling problems, are the simplest platform to implement models due to the rich variety of available tools. However, computers, limited by their generality, perform sub-optimally relative to custom hardware solutions. The goal of this thesis is to develop a new cost-effective and easy-to-use platform delivering orders of magnitude improvement in throughput over personal computers.
We suggest that FPGAs, or field programmable gate arrays, provide an outlet for dramatically enhanced performance. FPGAs are high-speed, reconfigurable devices that can implement any digital logic operation using an array of parallel computing elements. Already common in fields such as signal processing, radar, medical imaging, and consumer electronics, FPGAs have yet to gain traction in neural modeling due to their steep learning curve and lack of sufficient tools despite their high-performance capability. The overall objective of this work has been to overcome the shortfalls of FPGAs to enable adoption of FPGAs within the neural modeling community.
We embarked on an incremental process to develop an FPGA-based modeling environment. We first developed a prototype multi-compartment motoneuron model using a standard digital-design methodology. FPGAs at this point were shown to exceed software simulations by 10x to 100x. Next, we developed canonical modeling methodologies for manual generation of typical neural model topologies. We then developed a series of tools and techniques for analog interfacing, digital protocol processing, and real-time model tuning. This thesis culminates with the development of Dynamo, a fully-automated model compiler for the direct conversion of a model description into an FPGA implementation.
|
54 |
Techniques for FPGA neural modelingWeinstein, Randall Kenneth. January 2006 (has links)
Thesis (Ph.D)--Bioengineering, Georgia Institute of Technology, 2007. / Committee Chair: Lee, Robert; Committee Member: Butera, Robert; Committee Member: DeWeerth, Steve; Committee Member: Madisetti, Vijay; Committee Member: Voit, Eberhard. Part of the SMARTech Electronic Thesis and Dissertation Collection.
|
55 |
Diagnosis and error correction for a fault-tolerant arithmetic and logic unit for medical microprocessorsSavulimedu Veeravalli, Varadan. January 2008 (has links)
Thesis (M.S.)--Rutgers University, 2008. / "Graduate Program in Electrical and Computer Engineering." Includes bibliographical references (p. 91-96).
|
56 |
Implementation of adaptive digital FIR and reprogrammable mixed-signal filters using distributed arithmeticHuang, Walter. January 2009 (has links)
Thesis (Ph.D)--Electrical and Computer Engineering, Georgia Institute of Technology, 2010. / Committee Chair: Anderson, David V.; Committee Member: Ferri, Bonnie H.; Committee Member: Hasler, Paul E.; Committee Member: Kang, Sung Ha; Committee Member: McClellan, James H.; Committee Member: Wolf, Wayne H. Part of the SMARTech Electronic Thesis and Dissertation Collection.
|
57 |
Número do tipo ponto flutuante com precisão estendidaNunes, Richardson Leandro [UNESP] 31 March 2008 (has links) (PDF)
Made available in DSpace on 2014-06-11T19:23:38Z (GMT). No. of bitstreams: 0
Previous issue date: 2008-03-31Bitstream added on 2014-06-13T18:10:01Z : No. of bitstreams: 1
nunes_rl_me_ilha.pdf: 586695 bytes, checksum: b873e3bbf97a9b315f1c94485b17f2dc (MD5) / A execução de cálculos computaciona is é limit ada pela precisão que as linguagens de programação podem fornecer. Os compiladores possuem formatos de números com informação insuficiente para realizar cálculos que exijam grande precisão, porém, possuem ferramentas que possibilitam a criação de formatos extras. Utilizando o conceito de classe, é possíve l criar objetos computacionais e métodos. Visando solucionar problemas de precisão criou-se uma classe na qual o objeto é um número de ponto flutuante aqui chamado de Sfloat. A classe implementada em C++ é composta de um arranjo de variáveis booleanas de tamanho arbitrário para representar os bit s de um número de ponto flutuante e os métodos de classe para representar operadores aritmét icos e lógicos. Os operadores binários aritmét icos estão sobrecarregados, ou seja, os quatro operadores já existentes (“+”, “-”, “*” e “/”) podem ut ilizar números Sfloat como argumentos. Os operadores binários lógicos relacionais (<, >, <=, >=, ==, !=) seguem o mesmo modelo dos binários aritmét icos, sendo sobrecarregados para ut ilizar Sfloat como argumento. Para somar dois argumentos, soma-se dígito a dígito os dois argumentos. A subtração, na verdade, é a soma de um número posit ivo com um negat ivo, de modo que pode ser executada da mesma maneira que o operador soma, porém invertendo o sinal do segundo argumento. Na mult iplicação, somam-se as mult iplicações parciais de cada dígito de um dos fatores pelo outro fator, ou seja, a mult iplicação é executada como um somatório de mult iplicações parcia is. A divisão forma os dígitos do quociente verificando sempre qual o maior núme ro inteiro que pode mult iplicar o divisor sem ultrapassar o valor do dividendo. Sfloat foi ut ilizado para cálculos simples de soma, subtração... / Solut ion of computational problems involving float ing point numbers is limit ed by the accuracy that softwares can deliver. Most found compilers have usually float ing point kinds that are incapable to support the solut ion of numerical problems that need large accuracy on the final result or on the int ermediate comput ing steps. But such compilers are extensible allowing the development of derived data types and abstract data types and classes of high-accuracy numbers of float ing point kind. Using the c lass concept, it is possible to create computat ional objects and to implement methods (or member funct ions) owned by the object and that will act on the object data (or member variables). Aiming to solve numerical problems that happen in scient ific comput ing it was implemented a class to create float ing point numbers wit h high accuracy and range. That class was ca lled Sfloat. That class was implemented using a C++ compiler and is composed by an array of bits with variable size at compiler t ime. That extended float ing point number is based on the IEEE standard for float ing point numbers. The array bits carr y informat ion about signal (the first bit), exponent and mant issa. By varying it s array s ize it is possible to use hundreds or even thousands of bit s and so the mant issa precision can be very accurate with tens, hundreds or even thousands of decimal places, and also the exponent range can be very broad. Arithmet ic operators (+, -, *, /) can be overloaded in most compilers and also under C++ compilers. Overloading was used in this development by extending the meaning of the arithmet ic operators to allow its use also wit h Sfloat. Overloading process was also used to implement the relat ional logic operators. Nowadays, Sfloat can be used in implementat ions of most numerica l algorithms where is used arithmet... (Complete abstract click electronic access below)
|
58 |
Número do tipo ponto flutuante com precisão estendida /Nunes, Richardson Leandro. January 2008 (has links)
Orientador: João Batista Aparecido / Banca: Emanuel Rocha Woiski / Banca: Tito Dias Júnior / Resumo: A execução de cálculos computaciona is é limit ada pela precisão que as linguagens de programação podem fornecer. Os compiladores possuem formatos de números com informação insuficiente para realizar cálculos que exijam grande precisão, porém, possuem ferramentas que possibilitam a criação de formatos extras. Utilizando o conceito de classe, é possíve l criar objetos computacionais e métodos. Visando solucionar problemas de precisão criou-se uma classe na qual o objeto é um número de ponto flutuante aqui chamado de Sfloat. A classe implementada em C++ é composta de um arranjo de variáveis booleanas de tamanho arbitrário para representar os bit s de um número de ponto flutuante e os métodos de classe para representar operadores aritmét icos e lógicos. Os operadores binários aritmét icos estão sobrecarregados, ou seja, os quatro operadores já existentes ("+", "-", "*" e "/") podem ut ilizar números Sfloat como argumentos. Os operadores binários lógicos relacionais (<, >, <=, >=, ==, !=) seguem o mesmo modelo dos binários aritmét icos, sendo sobrecarregados para ut ilizar Sfloat como argumento. Para somar dois argumentos, soma-se dígito a dígito os dois argumentos. A subtração, na verdade, é a soma de um número posit ivo com um negat ivo, de modo que pode ser executada da mesma maneira que o operador soma, porém invertendo o sinal do segundo argumento. Na mult iplicação, somam-se as mult iplicações parciais de cada dígito de um dos fatores pelo outro fator, ou seja, a mult iplicação é executada como um somatório de mult iplicações parcia is. A divisão forma os dígitos do quociente verificando sempre qual o maior núme ro inteiro que pode mult iplicar o divisor sem ultrapassar o valor do dividendo. Sfloat foi ut ilizado para cálculos simples de soma, subtração... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Solut ion of computational problems involving float ing point numbers is limit ed by the accuracy that softwares can deliver. Most found compilers have usually float ing point kinds that are incapable to support the solut ion of numerical problems that need large accuracy on the final result or on the int ermediate comput ing steps. But such compilers are extensible allowing the development of derived data types and abstract data types and classes of high-accuracy numbers of float ing point kind. Using the c lass concept, it is possible to create computat ional objects and to implement methods (or member funct ions) owned by the object and that will act on the object data (or member variables). Aiming to solve numerical problems that happen in scient ific comput ing it was implemented a class to create float ing point numbers wit h high accuracy and range. That class was ca lled Sfloat. That class was implemented using a C++ compiler and is composed by an array of bits with variable size at compiler t ime. That extended float ing point number is based on the IEEE standard for float ing point numbers. The array bits carr y informat ion about signal (the first bit), exponent and mant issa. By varying it s array s ize it is possible to use hundreds or even thousands of bit s and so the mant issa precision can be very accurate with tens, hundreds or even thousands of decimal places, and also the exponent range can be very broad. Arithmet ic operators (+, -, *, /) can be overloaded in most compilers and also under C++ compilers. Overloading was used in this development by extending the meaning of the arithmet ic operators to allow its use also wit h Sfloat. Overloading process was also used to implement the relat ional logic operators. Nowadays, Sfloat can be used in implementat ions of most numerica l algorithms where is used arithmet... (Complete abstract click electronic access below) / Mestre
|
59 |
Software implementation of cryptography for wireless sensors and mobile processors = Implementação em software de criptografia para sensores sem fio e processadores móveis / Implementação em software de criptografia para sensores sem fio e processadores móveisGouvêa, Conrado Porto Lopes, 1984- 11 August 2013 (has links)
Orientador: Julio César López Hernández / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-24T00:18:34Z (GMT). No. of bitstreams: 1
Gouvea_ConradoPortoLopes_D.pdf: 3342900 bytes, checksum: b3a072bdbc369f1a1fffd95e1c1f4365 (MD5)
Previous issue date: 2013 / Resumo: A implementação eficiente e segura de esquemas criptográficos é um aspecto importante da criptografia aplicada. Neste trabalho, foca-se na implementação em software de algoritmos relevantes da criptografia de curvas elípticas (CCE), criptografia baseada em emparelhamentos (CBE), e de cifração autenticada (CA). Duas plataformas computacionais modernas foram utilizadas: o microcontrolador MSP430, bastante utilizado em redes de sensores sem fio, e o processador ARM, amplamente empregado por dispositivos móveis como smartphones e tablets que estão se tornando cada vez mais populares. Técnicas para a melhoria de desempenho em software utilizando conjuntos de instruções, periféricos e melhorias algorítmicas são descritas. A implementação segura, cujo objetivo é prevenir certos ataques de canais secundários, também é estudada e novas técnicas são providas para reduzir seu impacto na velocidade em processadores ARM. Tais resultados contribuem para a construção eficiente e segura de sistemas criptográficos em sensores sem fio e processadores móveis / Abstract: The efficient and secure implementation of cryptographic schemes is an important aspect of practical cryptography. In this work, we focus on the software implementation of relevant algorithms in elliptic curve cryptography (ECC), pairing-based cryptography (PBC) and in authenticated encryption (AE). Two modern computational platforms were targeted: the MSP430 microcontroller often used in wireless sensor networks, and the ARM processor, widely employed in mobile devices such as smartphones and tablets which are increasingly becoming ubiquitous. Techniques for improving the software performance by taking advantage of instruction sets, peripherals and algorithmic enhancements are described. The secure implementation, which aims at thwarting common side-channel attacks, is also studied and new techniques are provided for improving its efficiency on ARM processors. These results contribute to the building of efficient and secure cryptographic systems on wireless sensors and mobile processors / Doutorado / Ciência da Computação / Doutor em Ciência da Computação
|
60 |
Hardware Modules for Safe Integer and Floating-Point ArithmeticRatan, Amrita January 2013 (has links)
No description available.
|
Page generated in 0.0835 seconds