• No results found

Architecture. Evaluation and Classification of Computer Architectures. Bewertung und Klassifikation von Rechnerarchitekturen.

N/A
N/A
Protected

Academic year: 2021

Share "Architecture. Evaluation and Classification of Computer Architectures. Bewertung und Klassifikation von Rechnerarchitekturen."

Copied!
9
0
0

Loading.... (view fulltext now)

Full text

(1)

1

Evaluation and Classification of Computer

Evaluation and Classification of Computer

Architectures

Architectures

Bewertung und Klassifikation von

Bewertung und Klassifikation von

Rechnerarchitekturen

Rechnerarchitekturen

3 2

Architecture

" System Architecture " Software Architecture " Hardware Architecture

" Hardware Architecture = Computer Architecture " ISA: instruction set architecture

" HSA: hardware system architecture "

"

ISA: Hardware / Software Interface

Architect

searches the space of possible designs for the best solution (best design).

Good Ideas Good Ideas Mediocre Ideas Cost / Performance Analysis Goals

Criteria for a good design

Architectural Goals

High Performance

High Dependability

reliability, availability, maintainability, security, safety

Ease of Use

Extendability

Scalelability

(2)

5

Architektur-Kriterien / Criteria

" Modularität /modularity " Orthogonalität /orthognality " Angemessenheit /adeguacy " Symmetrie /symmetry " Virtualität / virtuality " Transparenz/ transparency Balance 6

Architectural Criteria

Modularity: Collection of (nearly) independent modules

extendability, availability

Orthogonality: Each module performs a well defined function; no two modules perform the same function

maintainability, costs, usage

Architectural Criteria

Adequacy: The costs of a module is adequate with respect to its function and use.

performance, costs

Symmetry: Existence of one module implies the existence of certain other modules

ease of use

Architectural Criteria

Virtuality: Elimination of physical limits (by software)

virtual processor, virtual memory, virtual channel, ....

scaleability, extendability, ease of use

Transparency: Hiding of 'unimportant' details from the user

(3)

9

Transparency

Transparency of access: identical operations for accessing different media ( main memory, mass storage, network,..)

ease of use

Transparency of location: access of objets without knowledge of their location

ease of use

Transparency of replication: transparent use of several copies of an object

performance, dependability

Transparency of concurrency: transparent use of parallelism performance, extendability, scaleability

10

Transparency

Transparency of performance: system configuration can be changed without loss of functionality, performance, reliability etc.

⇒ extenability, dependability, maintainability

Transparency of scale: system size can be extended such that its performance increases

⇒ scaleability, costs

Transparency of of faults: the systemhides, masks, tolerates faults and errors ⇒ dependability

Cost/Performance Analysis - Architektur-Analysen

Machbarkeitsstudie / Feasability Analysis

Leistungsbewertung / Performance Analysis

Verläßlichkeitsbewertung / Dependability Analysis

Kostenanalyse / Cost Analysis Performance

Dependability Costs

E-V-V

Verification / Verifikation: Correctness of the design

Is the system correctly designed and implemented?

Evaluation / Bewertung: What are the parameter values? What can the system deliver?

Validation / Validierung: Does the system what is expected?

Has the correct system been designed and implemented? - verification and evaluation of architectures - validation and evaluation of systems

(4)

13

Bewerten von System- und Rechnerarchitekturen

"

Ziele / Goals

"

Kriterien / Criteria

"

Bewertungsmaße / Measures

"

Bewertungsverfahren / Methods

Evaluation of Architectures

14

Evaluation of Architectures

Goals:

identification of weak points in the architecture identification of performance bottlenecks identification of hidden reserves

predictions of performance, reliability and costs comparison of architectures

evaluation of new concepts tuning for the application etc.

Leistungsgrößen / Performance Measures

" Raten / rates : MIPS, MFLOPS

" CPU - Performance oder

mittlerer CPU-Durchsatz (CPU-Throughput)

" Beschleunigung / Speed up

" Summarische Bewertung: Benchmarks

Zuverlässigkeitsgrößen / Dependability Measures

" Fehlerraten / failure rates : FITS

" Funktionswahrscheinlichkeit (Zuverlässigkeit)

/reliability

" Verfügbarkeit / Availability

" Sicherheit / Safety

(5)

17

Bewertungsverfahren / Means

" Messende Bewertung / Measurements

– Monitoring - Benchmarks

" Simulative Bewertung / Simulation " Analytische Bewertung / Analysis

" Wissensbasierte Bewertung / Knowledge based

Evaluation (Expertensysteme)

Simulative und analytische Bewertung sind modell-basiert.

18

Classification of Computer Architectures

Classification of Computer Architectures

Criteria / Klassifikations-Kriterien

Architecture Classes / Klassen

Criteria

"

Organization of the Control and Data Flow

"

Address Space Organization

"

Use of Physical Memory

Classification (1)

" Dependent Program Control:

– single, central control unit

" Independent Program Control:

(6)

21 e) Superscalar Processors C P U M e m o r y C P U M e m . M e m . M e m . M e m . M e m . M e m . C a c h e a . ) S I S D m o d e l b . ) m e m o r y i n t e r l e a v i n g c . ) w i t h c a c h e C P U M e m . M e m . M e m . C a c h e f1 f2 f n d . ) d . ) p i p e l i n e C P U w i t h n s t a g e s

Examples: Single Control Architectures

22

Classification (1)

" Dependent Program Control:

– single, central control unit

" Independent Program Control:

– parallel machines

At lower end performance increases faster than linearly, then the curve saturates and small gains in performance come at tremendous increase in cost. Performance Cost 1960s 1970s 1980s 1990s

Cost-performance curve of serial computers and its change over the decades

Cost-performance of serial computers Why Parallel Computing?

Although speed of computers is steadily increasing, new applications require even higher speed.

Performance of serial computers is beginning to saturate.

A natural way to solve the problem is the usage of an ensemble of processors

⇒ parallel computers

(7)

25 Faster solutions

Solution of large size problems

A parallel computer is simply a collection of processors, interconnected in a certain fashion to allow the coordination of their activities and the exchange of data

Parallel computers need parallel algorithms, i.e. algorithms suitable for implementation on parallel computers

Main Purposes for Parallel Computing

26

Flynn’s Classes

of

Computer

Architectures

" SI: Single Instruction Stream

" MI: Multiple Instruction Stream

" SD: Single Data Stream

" MD: Multiple Data Stream

Memory

CPU

Flynn’s Classes

of Computer Architectures

" SISD: Single Instruction Stream-Single Data Stream

– Monoprocessor

" Princeton (Von Neumann) -Architecture:

common instruction and data stream

" Harward - Architecture

separate instruction and data streams

Memory

CPU

Flynn’s Classes

of Computer Architectures

" SIMD: Single Instruction Stream-Multiple Data Stream •array processors, vector computers

CPU Memory

(8)

29

Flynn’s Classes

of Computer Architectures

" MIMD: Multiple Instruction Stream-Multiple Data

Stream

CPU Memory

30

Massively Parallel Processor: MPP

Hugh number of processing nodes Hugh switching network

Single computing resource - for a single job

Moderate Parallel Computers:

SMP (Symmetric Multiprocessor)

Degree of Parallelism

Symmetric Multiprocessor

In einem (symmetrischen) Multiprozessorsystem können alle Prozessoren auf einen gemeinsamen Hauptspeicher zugreifen. Dadurch ist es ihnen möglich, Daten auszutauschen.

Den Prozessoren steht aber nur ein gemeinsamer Datenweg zum Hauptspeicher zur Verfügung, z.B. ein gemeinsamer Bus.

Haupt- und Ein/Ausgabe-Prozessoren Hauptspeicher

(9)

33 Address Space P M Interconnection Network P M P M

Verbindungsnetz, interconnection network, interconnect Global Address Space

Distributed Physical Memory

34 Address Spaces P M Interconnection Network P M P M

Local Address Spaces

distributed and private physical memory

Parallel Architectures

" Control / Data Flow: centralized or distributed (Flynn)

" Address Space: local or global

" Physical Memory: private or shared

References

Related documents

“individuals are ‘always elsewhere’—always distracted or distractible from the social and physical presence of here and now to a more appealing attentional target” (Wallis

KNMI 24 h data KNMI 1 h data KNMI / DWD 5 min data >30 hourly automatic rain gauge measure- ments >300 manual rain gauge recordings radar data 1x1 km Municipality /

• Students in the Medical Assisting, Medical Laboratory Technician, and Associate’s Nursing programs must complete 50% of their program requirements at Rasmussen College, and

We conduct a systematic literature review and enrich it with interview-based knowledge from blockchain experts to show how blockchain technology changes business sectors, name

*It may be necessary to increase the trip cost insured on the travel protection plan for participants who require single accommodations, add the optional extension, and for those

El Niño (La Niña) years were defined as either: (a) type 1 ENSO events beginning in the second half of the previous year and ending in the first half of the same year as the

In an ACO algorithm, the colony of artificial ants is the stochastic construction procedure, which builds a complete solution probabilistically by iteratively adding solution

Typically, the optimiza- tion purpose is to set the optimal feeding trajectory, be- ing the feeding profile over time given by a piecewise lin- ear function, in order to reduce