Без темы
<<  Derivatives and Commodity Hedging Deutsch-russische Interessen identifizieren und zweiseitig finanzieren  >>
Target Platform
Target Platform
Why is Performance Analysis Difficult
Why is Performance Analysis Difficult
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Examples Event Stream Processing
Architecture Templates
Architecture Templates
Mapping Model
Mapping Model
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Optimization with conflicting goals
Multi-objective Optimization
Multi-objective Optimization
Multi-objective Optimization
Multi-objective Optimization
Multi-objective Optimization
Multi-objective Optimization
Evolutionary Algorithms
Evolutionary Algorithms
Evolutionary Algorithms
Evolutionary Algorithms
Evolutionary Algorithms
Evolutionary Algorithms
Evolutionary Algorithms
Evolutionary Algorithms
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
An Evolutionary Algorithm in Action
EXPO – Tool architecture
EXPO – Tool architecture
EXPO – Tool architecture
EXPO – Tool architecture
PISA: Implementation
PISA: Implementation
EXPO - Tool
EXPO - Tool
EXPO - Tool
EXPO - Tool
Results
Results
Картинки из презентации «Design Space Exploration of Embedded Systems» к уроку английского языка на тему «Без темы»

Автор: . Чтобы познакомиться с картинкой полного размера, нажмите на её эскиз. Чтобы можно было использовать все картинки для урока английского языка, скачайте бесплатно презентацию «Design Space Exploration of Embedded Systems.ppt» со всеми картинками в zip-архиве размером 2543 КБ.

Design Space Exploration of Embedded Systems

содержание презентации «Design Space Exploration of Embedded Systems.ppt»
Сл Текст Сл Текст
1Design Space Exploration of Embedded 22task4. risc.
Systems. © Lothar Thiele ETH Zurich. 23Abstraction. Idea: unified view of
2Overview. Review of General Aspects task scheduling, arbitration and event
Basic Models and Methods Multi-Criteria scheduling in networks: methods: queueing
Optimization Applications. theory (statistical bounds, markov chains)
3Target Platform. Heterogeneous real-time calculus (worst case bounds,
computing and memory resources some min-max algebra).
resource types: GP processors, ASIPs (DSP, 24Real-Time Calculus. Example of a
micro-controller), weakly programmable dynamic analytic model Characteristics
co-processors, re-configurable components, yields worst case estimation results for
hard coded IP components heterogeneous memory, delay, throughput takes into
platform software: RTOS, scheduling account application structure (task graph
(pre-emptive, static, dynamic), representation) architecture and mapping
synchronization. RISC. image coprocessor. (computation, communication, scheduling)
SDRAM. DSP. CAN interface. mC. FPGA. environment (characterization of input
4Target Platform. Communication traces).
micro-network on chip for synchronization 25Elements of Modular Performance
and data exchange consisting of busses, Analysis. application. hardware
routers, drivers some critical issues: architecture. mapping, scheduling. system
topology, switching strategies (packet, architecture model. performance model.
circuit), routing strategies (static – architectural element model. environment
reconfigurable – dynamic), arbitration model. analysis.
policies (dynamic, TDM, CDMA, fixed 26Application. Application. p=1 s, j=0.2
priority) challenges: heterogeneous s. Hndl. Dec. Disp.
components and requirements, compose 27Hardware Architecture. Application. HW
network that matches the traffic Architecture. p=1 s, j=0.2 s. 22 MIPS. 10
characteristics of a given application MIPS. 72 kbps. Hndl. Dec. Disp.
(domain). 28Mapping. Application. Mapping. HW
5Design Space. LookUp. RISC. EDF. mE. Architecture. p=1 s, j=0.2 s. 22 MIPS. 10
mE. mE. TDMA. static. Priority. mE. mE. MIPS. 72 kbps. Hndl. Dec. Disp.
mE. WFQ. Cipher. DSP. 29Modular Performance Analysis.
6Design Space Exploration. application. hardware architecture.
7Design Space Exploration. mapping, scheduling. system architecture
8Design Space Exploration. (Semi-) model. performance model. architectural
Automated Design Space Exploration. element model. environment model.
9Why is Performance Analysis Difficult? analysis.
task1. task2. task3. task4. complex 30Overview. Review of General Aspects
behavior - input stream - data dependent Basic Models and Methods Multi-Criteria
behavior. interference - limited resources Optimization Applications.
- scheduling/arbitration. interference of 31Why Performance Analysis? (Semi-)
multiple applications - limited resources Automated Design Space Exploration.
- scheduling/arbitration - anomalies. 32Optimization with conflicting goals.
CPU1. I/O. CPU2. DSP. Multiobjective optimization: Find a set of
10Simulation. Target architecture optimal trade-offs Example: computer
co-simulation combines functional and design.
performance validation extensive runtimes 33Multi-objective Optimization.
worst case inputs ? test case definition ? 34Multiobjective Optimization. Maximize
re-targeting expensive. mixed model (y1, y2, …, yk) = ?(x1, x2, …, xn) Pareto
function: application structure: set = set of all Pareto-optimal solutions.
hardware-software architecture. input 35Multiobjective Optimization. Minimize.
trace. output trace. (x1, x2, …, xn) (y1, y2, …, yk)
11Trace-Based Simulation. Steps: Difficulties: ? large search space ?
execution trace determined by multiple optima. f. x2. y2. objective
co-simulation abstract representation space. decision space. dominated. Pareto
using communication graph extension of optimal = not dominated. y1. x1.
graph by actual architecture Faster than 36Optimization Alternatives. Use of
simulation, but still based on single classical single objective optimization
trace. input trace. functional model. methods simulated annealing, tabu search
complete trace. abstract graph. integer linear program other constructive
communication architecture. trace or iterative heuristic methods Decision
simulation. estimation results. [Lahiri making (weighting the different
et. al, 2001]. objectives) is done before the
12Static Analytic Models. Steps: optimization. Population based
describe computing, communication and optimization methods evolutionary
memory resources by algebraic equations, algorithms genetic algorithms Decision
e.g. describe properties of input using making is done after the optimization.
parameters, e.g. input data rate combine 37Traditional Approaches. Example:
relations Fast and simple estimation weighting approach. parameters. multiple
Generally inaccurate modeling of shared objectives. single objective.
resources. transformation. y. (y1, y2, …, yk).
13Dynamic Analytic Models. Combination 38Evolutionary Algorithms. Principles of
between static models, possibly extended Evolution. ? Cross-over. ? Selection. ?
by their dynamic behavior, e.g. Mutation.
non-determinism in run-time and event 39A Generic Multiobjective EA.
processing dynamic models for describing population. archive. evaluate sample vary.
shared resources (scheduling and update truncate. new archive. new
arbitration) Variants queuing theory population.
(statistical models, average case) 40An Evolutionary Algorithm in Action.
real-time calculus (interval methods, hypothetical trade-off front.
worst case) More accurate than static 41Overview. Review of General Aspects
models. Basic Models and Methods Multi-Criteria
14Dynamic Analytic Models. component Optimization Applications.
simulation. data sheets. model of 42EXPO – Tool architecture. MOSES. EXPO.
components. input traces. model of SPEA 2. Tool available online:
environment. system model. model of http://www.tik.ee.ethz.ch/expo/expo.html.
architecture. spec. of inputs. estimation Exploration Cycle.
results. 43The Concept of PISA. Algorithms
15Summary. Timing Accuracy. Simulation. Applications. SPEA2. knapsack. NSGA-II.
Trace-based simulation. Dynamic analytic TSP. network processor design. PAES.
methods. Static analytic methods. Platform and programming language
Run-time. independent Interface for Search
16Bounds, Guarantees and Predictability. Algorithms [Bleuler et al.: 2002].
Example: end-to-end delay. design. 44PISA: Implementation. shared file
17Overview. Review of General Aspects system. selector process. variator
Basic Models and Methods Multi-Criteria process. text files. application
Optimization Applications. independent: mating / environmental
18Examples Event Stream Processing. selection individuals are described by IDs
19Application Model. Example of a simple and objective vectors. handshake protocol:
stream processing task structure: state / action individual IDs objective
20Architecture Templates. In general, we vectors parameters. application dependent:
assume an arbitrary heterogeneous variation operators stores and manages
architecture consisting of computing individuals.
resources, memory and communication 45PISA Website.
resources. event. event. events. http://www.tik.ee.ethz.ch/pisa.
21Mapping Model. 46EXPO - Tool.
22Allocation and Binding. Allocation can 47Results. Performance for
be represented as a function: Binding is a encryption/decryption. Performance for RT
relation: Binding restrictions: task1. voice processing.
class. task2. filter. task3. schedule. 48Validation Strategy.
Design Space Exploration of Embedded Systems.ppt
http://900igr.net/kartinka/anglijskij-jazyk/design-space-exploration-of-embedded-systems-121000.html
cсылка на страницу

Design Space Exploration of Embedded Systems

другие презентации на тему «Design Space Exploration of Embedded Systems»

«Periodic table of elements» - 2Co + O2 = 2CoO. you have viewed the presentation of some chemical elements. inert gas. In the absence of cobalt akobaltoz develops. 1751 swedish mineralogist Cronstedt. Co + 2H2SO4 = CoSO4 + SO2 + 2H2O. Density (at n. Y.) = 8.902 g / cm ? Melting point = 1726 K. It is a part of vitamin B12 (cobalamin).

«Современное обучение английскому» - Мягкая обложка 144 стр., 210x285 мм. Английский язык. Английский язык – 2011». Курс «Happy English.ru». “Read up! / Почитай!” для начальной школы. Новый проект - серия книг для чтения. Методика обучения иностранным языкам: традиции и современность под редакцией А.А. Миролюбова. Публикации учителей в учебно-методическом журнале “Английский язык в школе” (“АЯШ”).

«Animals names» - c...w. Animals. be...r. b...rds. ...heep. h...re. Тренажёр «Вставь пропущенную букву". p...ppy. co...k.

«Еда на английском» - Если ответ правильный по ссылке включается слайд 5. Правильные ответы скрыты с помощью затенения экрана. Обобщающий урок по теме «Еда». Пояснительная записка. Презентация Smart Notebook прилагается. Способы работы со слайдами описаны с помощью сносок. Возможность передвигать слова делает более интересным простое задание соотнести слова и картинки.

«Переводчик с английского на русский» - Группа STAR была основана в 1984г. в городе Штейн-на-Рейне в Швейцарии. Выравнивание памяти переводов. Минусы: большие затраты на вычитку. Минусы: поддержка памяти переводов. Терминологическая проверка. Орфографическая и грамматическая проверки. Группа STAR. Плюсы: минимальная компьютерная подготовка переводчика.

«Артикли в английском языке» - The bear is smiling! The flower. Артикль не употребляется перед именами и фамилиями. Запомни! A fish. Например: Ivan Vladimirovich. Вставь артикли a/an или the, если необходимо. … man gives little … Ivan Vladimirovich. Definite and indefinite article. English humor. The Black Sea. Артикли бывают: Неопределенные a/an определенные the.

Без темы

661 презентация
Урок

Английский язык

29 тем
Картинки