Basic Computer Structure and
Knowledge
For this purpose we do not need a description of all the details and design features of modern computers. Just a general lay-out of the very basics is necessary. I will devote much attention to the general workings of Boolean logic (hardware) circuits (but only a very simple example of them will be treated of here [Part Two of this Essay]). Thereby it is important to realize that the computer-hardware is a physical device that obeys the laws of physics. Further one must realize that a computer can simulate phenomena of the outside world, and although these simulations are not the same as the phenomena which are being simulated, those simulations are something in their own right. What they are (in their own right) depends on their general and detailed structure and/or behavior. In the next Essay (about artificial life) these subjects will be adressed fully.
In Part One of this Essay I will treat of the general CONCEPT of a digital computer, in terms of Turing machines. A concept already worked out in the 19-thirties.
There are two main types of computers, analog computers and digital computers. Digital computers are discrete machines having access to a finite number of internal states only, while analog machines have access, in principle, to an infinite number of internal states and could therefore be expected to outperform digital machines. Examples of this enhanced ability of analog computers would include solving halting-problems (i.e. in principle be able to determine in advance whether any program, fed into the computer, will or will not yield a definite result in a finite number of steps), and generating non-computable numbers. For a digital computer one cannot write a test-program that could determine for any other program, to be run on such a computer, whether that program (to be tested) will, when run, give a definite result after a finite number of steps. Also a digital computer cannot compute certain numbers (so called non-computable numbers).
Further we have serial machines and parallel machines. While a serial machine can only perform one computational step after another, a parallel machine can execute more than one such steps simultaneously. Parallel machines accordingly consist of more than one processor, which operate in harmony. Many processes in Nature are in fact proceeding in a parallel fashion, and can therefore adequately be simulated by such machines. But it is possible to simulate such a parallel machine on a serial machine.
In our discussions and explanations we will confine ourselves mainly to DIGITAL SERIAL COMPUTING MACHINES, but the principle (concept) expounded covers parallel machines as well.
Figure 1. Von Neumann's computer architecture -- the layout of a typical serial machine. Except with respect to input and output devices, the information-flow is in a back-and-forth fashion.
To program such a computer, in order that it will solve a certain problem or generate some desired result, the programmer first writes an algorithm, which is a solution of the problem in the form of a sequence of steps, written in ordinary language. This algorithm is then coded into a suitable programming language that will enable the computer to ' understand ' and successfully execute the corresponding instructions. Usually, this involves one of the ' higher ' programming languages, so called because they are reasonable close to human language. But because the Central Processing Unit (CPU) is composed of a set of Boolean logic circuits, it is capable only of performing elementary arithmetic operations such as addition, subtraction, multiplication, and so on. Thus the original programming language fed into the computer must be first converted by means of an interpreter (translates one line of code and executes it, then translates the next line, etc.), or a compiler (translates the whole program and then executes it) into a machine-readable assembly language (instructions, coded in this assembly language are then directly translated into machine-code, that consists of the electronical equivalents of 0's and 1's) . Only then the machine is able to execute the fed-in instructions.
The input is coded up in memory, which is a grid of electronic on-off switches. The processor, which is a chip of integrated circuitry, alters what is in the memory, resulting in a different on-off pattern of switches, and then the output decodes and displays the new contents of the memory. So the actual computation consists of the processor's activities on the memory. Accordingly the processor and the memory stand in mutual contact with each other.
Virtually all computers, from home desktop machines to giant research
systems, all share a basic structure. They have a central processor that
computes and performs logic, memory that holds data, and input-output
devices.
Knowledge
The 5 Component Model
A computer system can be divided into 5 components:- Hardware
- Software
- Data
- Procedures
- Personnel
Introduction
This Essay is meant as a preparation for the next Essay, which is about the ontological status of ARTIFICIAL LIFE (a-life), mainly that brand of a-life that is computer-generated. For an assessment of the ontological status of the creations of a-life it is paramount to know something about the general and basic structure of a computer, in our case a digital computer. Only after possessing some insights into the structure of these interesting machines can we rationally speak about the role and nature of the substrate of artificial-life creatures in order to assess the reality-status of those creatures.For this purpose we do not need a description of all the details and design features of modern computers. Just a general lay-out of the very basics is necessary. I will devote much attention to the general workings of Boolean logic (hardware) circuits (but only a very simple example of them will be treated of here [Part Two of this Essay]). Thereby it is important to realize that the computer-hardware is a physical device that obeys the laws of physics. Further one must realize that a computer can simulate phenomena of the outside world, and although these simulations are not the same as the phenomena which are being simulated, those simulations are something in their own right. What they are (in their own right) depends on their general and detailed structure and/or behavior. In the next Essay (about artificial life) these subjects will be adressed fully.
In Part One of this Essay I will treat of the general CONCEPT of a digital computer, in terms of Turing machines. A concept already worked out in the 19-thirties.
There are two main types of computers, analog computers and digital computers. Digital computers are discrete machines having access to a finite number of internal states only, while analog machines have access, in principle, to an infinite number of internal states and could therefore be expected to outperform digital machines. Examples of this enhanced ability of analog computers would include solving halting-problems (i.e. in principle be able to determine in advance whether any program, fed into the computer, will or will not yield a definite result in a finite number of steps), and generating non-computable numbers. For a digital computer one cannot write a test-program that could determine for any other program, to be run on such a computer, whether that program (to be tested) will, when run, give a definite result after a finite number of steps. Also a digital computer cannot compute certain numbers (so called non-computable numbers).
Further we have serial machines and parallel machines. While a serial machine can only perform one computational step after another, a parallel machine can execute more than one such steps simultaneously. Parallel machines accordingly consist of more than one processor, which operate in harmony. Many processes in Nature are in fact proceeding in a parallel fashion, and can therefore adequately be simulated by such machines. But it is possible to simulate such a parallel machine on a serial machine.
In our discussions and explanations we will confine ourselves mainly to DIGITAL SERIAL COMPUTING MACHINES, but the principle (concept) expounded covers parallel machines as well.
The Digital (serial) Computer
The main components of a digital computer are :- Input devices (keyboard, mouse, etc).
- Memory board.
- Central Processing Unit (processor).
- Output devices (video terminal, printer, etc.).
Figure 1. Von Neumann's computer architecture -- the layout of a typical serial machine. Except with respect to input and output devices, the information-flow is in a back-and-forth fashion.
To program such a computer, in order that it will solve a certain problem or generate some desired result, the programmer first writes an algorithm, which is a solution of the problem in the form of a sequence of steps, written in ordinary language. This algorithm is then coded into a suitable programming language that will enable the computer to ' understand ' and successfully execute the corresponding instructions. Usually, this involves one of the ' higher ' programming languages, so called because they are reasonable close to human language. But because the Central Processing Unit (CPU) is composed of a set of Boolean logic circuits, it is capable only of performing elementary arithmetic operations such as addition, subtraction, multiplication, and so on. Thus the original programming language fed into the computer must be first converted by means of an interpreter (translates one line of code and executes it, then translates the next line, etc.), or a compiler (translates the whole program and then executes it) into a machine-readable assembly language (instructions, coded in this assembly language are then directly translated into machine-code, that consists of the electronical equivalents of 0's and 1's) . Only then the machine is able to execute the fed-in instructions.
The input is coded up in memory, which is a grid of electronic on-off switches. The processor, which is a chip of integrated circuitry, alters what is in the memory, resulting in a different on-off pattern of switches, and then the output decodes and displays the new contents of the memory. So the actual computation consists of the processor's activities on the memory. Accordingly the processor and the memory stand in mutual contact with each other.
Virtually all computers, from home desktop machines to giant research
systems, all share a basic structure. They have a central processor that
computes and performs logic, memory that holds data, and input-output
devices.
No comments:
Post a Comment