CMSI 284: Welcome to Week 02

This Week's Agenda

For this week, here's the plan, Fran…

Computer Systems Organization

In this course you will become familiar with all the parts that make up a computer. Of course, there is the Central Processing Unit or CPU, but there are many more parts in the box than just that as you know. We'll learn about many of them and their functions, both individually and as part of the system.

Part of any computer system is the input and output hardware which allows the computer system to actually be a part of the real world. For example, you need a keyboard for input and a display for output. You may need a printer for hard-copy. You DEFINITELY need a network card or device — otherwise you can't access the Internet — and you need ports so you can connect peripheral devices as well.

We will also look INSIDE the CPU to see how things function, and how this amazing device can do so many things so well. We'll deconstruct if from its highest view to its lowest level to see just what's going on under the hood.

You can view the computer system as a set of layers, with each layer working on the actions of the one beneath it. Here's an example diagram:

As can be seen in this version, a system can be viewed to have seven layers. Granted this is a simplified view, but the idea is apparent. Each layer builds upon the functionality that is in the layer below it, and provides functionality to the layer above it.

In this diagram you can see at each level the type of function that is provided as well as some of the typical things that are used to implement those functions. For example, at the lowest level is the actual digital logic, the hardware that all the operations of the computer run on. That is implemented by gates, circuitry, motherboards, daughterboards, chips or integrated circuits, and other hardware parts.

In the layer above that, the control layer, uses the hardware and contains the microcode which uses the hardware from the layer below it. Note that microcode can be done using internal software [also sometimes known as firmware], or it can be realized using hardwiring of components in the system.

In this course, we will be learning in some detail about the digital logic, control, machine, and assembly language layers, and will use both the assembly language and high-level language layers to facilitate that investigation with real-life examples.

Discussion: Are there any other parts to a computing system? What do you think they are? Would you consider them to be part of one of the system layers or something separate? Where does a human user of the system fit ~ is she part of the top layer of the model or is she on top of the entire model?

Von Neumann Architecture and Layers of Organization

Those of you who have taken CMSI 185, CMSI 186, and CMSI 281 have no doubt heard the name of John Von Neumann, the guy who is credited with the concepts of the modern computer and the computer architecture that bears his name. In those other classes, I have reduced the idea to three steps, which is a VERY simplified way of looking at it.

Quick Quiz: Does anyone remember what those three steps are?

In the diagram you can see the basic parts of a computer system. Each block in the diagram shows a collection of major components, and each of those collections has a HUGE number of components, literally thousands of them. These parts comprise what is known as a stored-program computer.

This architecture provides the following things:

  • The three main parts shown in the diagram
  • A single path between main memory and the control unit, which causes the machine to have to alternate between the instruction and execution operations
  • The capacity to carry out sequential instruction processing [note that in the case of stored instructions, this capacity is required by that alternation]

Briefly, the architecture runs programs using the following steps:

  • The control unit fetches the next program instruction from memory; the instruction's address is kept in an internal register known as the program counter.
  • The fetched instruction is decoded so that the processor's microcode can understand just what the instruction is supposed to do. Note that this may include one or more additional fetches from main memory to acquire the data that instruction might need; these data are then placed into registers in the CPU.
  • Finally the Arithmetic Logic Unit [ALU] executes the decoded instruction. Results may end up in another register, or may end up being placed back in main memory.

Another, expanded, view of the system looks like the diagram at the right. The same components are shown, but with the addition of a system bus which has three sets of connections between the main components.

The data bus is used to carry the bits that comprise data that need to be transferred between the CPU and main memory or the CPU and the input/output controller. This data is accessed using…

The address bus which is used by the CPU to access other parts of the system. The CPU will place an address consisting of a set of bits, onto the address bus to let the system know where the data it needs is expected to be.

Finally, the control bus lets the various parts of the system interact in a controlled way, by raising or lowering specific lines on the bus.

The fact that the control bus implements alternating between the instruction and execution operations causes what is known as the Von Neumann Bottleneck.

By the way, Von Neumann is know for a LOT more than just the stored-program computer architecture. Here are some other things he was involved in or responsible for:

  • set theory, including the Von Neumann Paradox
  • operator theory
  • measure theory
  • quantum logic
  • game theory
  • linear programming
  • fluid dynamics
  • Manhattan project and Atomic Energy Commission [AEC]
  • Cellular automata, DNA and the universal constructor
  • Weather systems and global warming
  • Technological singularity hypothesis

Devices and Peripherals and Software ~ More of the System

In addition to the basic parts as outlined above, there are many other things that connect to the system to help it function in a meaningful way. These are collectively referred to as devices and they come in three basic flavors, as you can see in the following table:

Input DevicesOutput DevicesStorage Devices
Light Pen
Track Pad
Voice Recognizer
Fingerprint Scanner
Card Reader
Game Controller
Data Glove
Video Camera
Eye Tacker
Motion Sensor
Printer (2D or 3D)
Film Recorder
Hologram Generator
Robot Arm
Voice Synthesizer
Card Punch
Disk Drive
CD Drive
DVD Drive
USB Flash Drive
Solid State Drive (SSD)
Tape Drive

Quick Quiz: Can you think of others? What have I missed here?

Software can be though of as split between Applications software and Systems software, as shown in the following table:

Applications SoftwareSystems Software
Written for peopleWritten for computers
Deals with human-centered abstractions like customers, products, orders, employees, players, users Deals with computer-centered concepts like registers and memory locations
Solves problems of interest to humans, usually in application areas like health care, game playing, finance, etc. Controls and manages computer systems
Concerned with anything high-level Concerned with data transfer, reading from and writing to files, compiling, linking, loading, starting and stopping programs, and even fiddling with the individual bits of a small word of memory
Is almost always device or platform independent; programs concentrate on general-purpose algorithms Deals with writing device drivers and operating systems, or at least directly using them; programmers exploit this low-level knowledge
Is often done in languages like JavaScript, Perl, Python, Ruby, Lisp, Elm, Java, and C# that feature automatic garbage collection and free the programmer from low-level worries Is often done in assembly language, C, C++, and Rust where programmers have to manage memory themselves
Is done in languages that generally have big fat runtime systems Generally feature extremely small run-time images, because they often have to run in resource constrained environments
If done properly, can be very efficient: good garbage collection schemes allow much more efficient memory utilization than the usual memory micro-management common in C programs If done properly, can be very efficient: you can take advantage of the hardware

Also in terms of software, there are different levels of programming languages: High-level languages, Assembly Languages, and Machine Languages.

Often, people write in a high-level language, then run that source code through a compiler which translates the code into assembly language, which an assembler then translates into machine language:

Here's an example. Start with this C++ function:

   long example(long x, long y, long z) {
       if (x > y) {
           return x * y - z;
       } else {
           return  (z * y) * y;

The compiler produces this assembly language:

           cmp     rdi, rsi
           jg      .L5
           imul    rdx, rsi
           mov     rax, rdx
           imul    rax, rsi
           mov     rax, rdi
           imul    rax, rsi
           sub     rax, rdx

which becomes this in machine language:


Discussion: Which language level would you rather program in, and why?

Processor Parts and Organization

As software developers, one important fact we ALWAYS need to remember is that ANYTHING THAT CAN BE DONE WITH SOFTWARE CAN ALSO BE DONE IN HARDWARE!. Why, then, isn't EVERYTHING done in hardware?

Quick Quiz: Any guesses?

The fact is that hardware runs faster, is more consistent in its operation, and is less prone to systemic errors in most cases than software. This makes it a prime candidate for implementing all kinds of things that could be done in software.

However, it takes more time and much more engineering rigor to develop hardware, and it is often orders of magnitude more expensive to correct mistakes. If a mistake is made in hardware, there are a number of ways to fix it, ranging from simple parts substitution all the way up to redesigning and remanufacturing a circuit board. With software, in most cases a developer can make a few simple changes to the code and rebuild the program quickly and easily.

Here is an article that illustrates this point; if you look about halfway down the page you'll see a description of some of the steps involved in creating the hardware for the F-14 Tomcat Fighter Plane [the plane used in the movie Top Gun].

But because of the stored program computing approach, the two rely on each other for proper operation. Hardware relies on its microcode, which is hard-wired into the CPU's circuitry, to interpret and execute its operations, called the instruction set. Each of these basic operations is represented by a specific pattern of bits which is known as the machine language opcode or simply opcode. A complete opcode will also contain patterns of bits which specify the data for the arguments that the instruction needs. Stored program computers executing software have to read these instructions from memory; they are placed there by the operating system when the application is started [loaded]. We've seen already how compiling the programs source code leads to the binary bit patterns that become these instructions.

There are several parts inside the CPU that are involved in the processing:

Assembly Language

Programmers are better off for learning assembly than they are not learning it. For one thing, it gives you a more intimate and hands-on study of computer systems as you can see from the following quote:

Assembly is a mechanism by which a programmer can learn details of computer hardware, CPU components, memory organization, and the interactions among these elements of computer architecture.

— Brian Hall and Kevin Slonka

Other reasons:

Topics in Computer Systems

Thanks to Dr. Toal, here are some things to study to get both breadth and depth about computer systems and systems programming, both concepts and real-world examples:

  • Digital Logic
  • How CPUs work (fetch-execute cycle, context-switching, memory segments)
  • The organization and components of a computer system
  • Memory, caches, and storage devices
  • Assembly and Machine Languages
  • The C Programming Language (and others like it)
  • Compilation, assembly, linking, loading, and execution
  • Basic operating system functionality (file, process, and memory management)
  • Interfacing with the O.S.
  • System-level I/O
  • Compiling, assembling, linking, loading
  • Optimizing programs

  • Exploiting systems
  • Defending systems from attack
  • File formats
  • Advanced operating system functionality (virtual memory, interrupts, bootstrapping)
  • Device Drivers
  • Concurrency basics: threads and events
  • Networking basics

Beginning Analysis of Number Systems

In order to start understanding how this all works, we need to start with the way computers store, handle, and recognize things, and how they can represent these things to their human counterparts. This means, first of all, we need to understand how computers deal with numbers.

So first of all, numbers are just symbols. The numbers we are used to, zero through nine, are what is known as the decimal system of course. It is based on the idea of groups of ten, probably because we have ten fingers. Interestingly, the idea of zero is a relatively new concept. We'll see more about that next week.

We can use combinations of any of these digits to represent any numerical value. Each successive number place to the left of the one we're looking at is exactly one order of magnitude [one power of ten] greater that the one immediately to its right. You know them as the one's place, ten's place, hundred's place, and so on.

For now, though we need to be aware that decimal or base 10 is different from the way a computer handles numbers. Digital computers are based on the binary system, which only has two states: ON and OFF, which we replace with one and zero. Don't worry if it seems abstract and difficult to understand how a computer can use only two symbols and do all of the fantastic things it can do. What's important is that you are aware of the idea of the place values of the binary system.

Where in the decimal system, each decimal digit can have any of ten values 0 – 9, each binary digit [which we contract to call a bit] can have only one of TWO values, 0 – 1. But like the decimal system digits, each of the number place values is based on a value of TWO rather than TEN.

So, looking at the number 1234 for example, we have a 4 in the one's place, a 3 in the ten's place, a 2 in the hundred's place, and a 1 in the thousand's place. Remember, though, that these are all powers of ten. For example, the ten's place is 101 and the hundred's place is 102. If we view the number in terms of powers of ten, we get something like:

   1 x 103 + 2 x 102 + 3 x 101 + 4 x 100

Another way to show this would be:

1000's100's 10's1's
103102 101100

Notice that as we move from right to left, the exponents keep increasing by one. This is important.

Now, let's turn to the binary system. In this case, the number BASE is different, so instead of powers of TEN, we have powers of TWO. Let's expand the table and replace the number base to see what it would look like in that system:

one-twenty-eight'ssixty-fours' thrity-two'ssixteen's eight'sfour's two'sone's
2726 2524 2322 2120

So now for a simple example. Let's take an 8-bit value [which is known as a byte] as follows:

one-twenty-eight'ssixty-fours' thrity-two'ssixteen's eight'sfour's two'sone's
2726 2524 2322 2120

If we do the math, we'll find we have:

   (1 x 27) + (1 x 25) + (1 x 24) + (1 x 22) + (1 x 20)

…which is:

   128 + 32 + 16 + 4 + 1 = 18110

In-class Assignment #2 — Running the Bases

Now, in class, get with your partner and work out the following for practice. The better you get at the ability to translate between bases, the easier it will be to understand the different ways that numbers are represented in the computer. Next week we'll start out with an easy way to convert from decimal to binary, but for now see if you can figure out a way on your own. Feel free to use the Internet for any assistance you can get. Here are the problems:

Homework Assignment #2

Reminder: homework02 is due at the end of THIS WEEK. All homework assignments are available from the syllabus page and also from the assignments page, but just to make sure you remember …

Week Two Wrap-up

That's probably enough for the this week. Be sure to check out the links to the related materials that are listed on the class links page.