2 CourseThis course aims to teach the fundamentals of computer organization and architecture through assembly language programming.Topics to be learned:Basic Computer Organization;Assembly Language.
3 Computer Architecture and Organization Architecture is concerned with attributes visible to the programmer:Instruction set, number of bits used for data representation, I/O mechanisms, addressing techniques.Organization is how features are implemented:Control signals, interfaces, memory technology.
4 Computer architecture Computer architecture deals with the functional behavior of a computer system as viewed by a programmer (like the size of a data type – 32 bits to an integer).All Intel x86 family share the same basic architecture.
5 Computer organization Computer organization deals with structural relationships that are not visible to the programmer (like clock frequency or the size of the physical memory).Organization differs between different versions.
6 The Organization of a Computer Processor:Data path - to do arithmetic and logic: e.g. adders, multipliers, shifters, … .Control - to give directions to the other components: e.g. tells the data path, memory, and IO devices what to do according to the instructions of the program.Memory: Holds data and instructions: e.g. cache, main memory, disk, … .Input: Sends data to the computer: e.g. keyboard, mouse, … .Output: Gets data from the computer: e.g. screen, sound card, printer, … .All computers have had 5 main components:ProcessorControlInputMemoryDatapathOutput
7 The Von Neumann ModelThe von Neumann model consists of five major components:(1) input unit;(2) output unit;(3) arithmetic logic unit;(4) memory unit;(5) control unit.
8 The Von Neumann Architecture Named after John von Neumann,Princeton, he designed acomputer architecture wherebydata and instructions would beretrieved from memory,operated on by an ALU, andmoved back to memory (or I/O)This architecture is the basis formost modern computers (onlyparallel processors and a fewother unique architectures usea different model).Hardware consists of 3 units:CPU (control unit, ALU, registers);Memory (stores programs and data);I/O System (including secondary storage).Instructions in memory are executed sequentially unlessa program instruction explicitly changes the order.
9 The von Neumann Model Instruction Cycle This is a general depiction of a von Neumann system:These computers employ a fetch-decode-execute cycle to run programs as follows . . .
10 The von Neumann Model: fetch The control unit fetches the next instruction from memory using the program counter to determine where the instruction is located.
11 The von Neumann Model: Decode The instruction is decoded into a language that the ALU can understand.
12 The von Neumann Model: Execute Any data operands required to execute the instruction are fetched from memory and placed into registers within the CPU.
13 The von Neumann Model: Execute The ALU executes the instruction and places results in registers or memory.
14 Non-Von Neumann Models Conventional stored-program computers have undergone many incremental improvements over the years.These improvements include adding specialized buses, floating-point units, and cache memories, to name only a few.But enormous improvements in computational power require departure from the classic von Neumann architecture.Adding processors is one approach.
15 Non-von Neumann Models In the late 1960s, high-performance computer systems were equipped with dual processors to increase computational throughput.In the 1970s supercomputer systems were introduced with 32 processors.Supercomputers with 1,000 processors were built in the 1980s.In 1999, IBM announced its Blue Gene system containing over 1 million processors.
16 Non-von Neumann Models Parallel processing is only one method of providing increased computational power.DNA computers, quantum computers, and dataflow systems. At this point, it is unclear whether any of these systems will provide the basis for the next generation of computers.
17 Non-von Neumann Models Leonard Adleman is often called the inventor of DNA computers. His article in a 1994 issue of the journal Science outlined how to use DNA to solve a well-known mathematical problem, called the "traveling salesman" problem. The goal of the problem is to find the shortest route between a number of cities, going through each city only once. As you add more cities to the problem, the problem becomes more difficult. Adleman chose to find the shortest route between seven cities. DNA computing is still in its infancy.
18 Harvard Architecture Data memory is separate from Program memory. Allows data and program information to move separately and so is faster than von Neumann.Harvard architecture uses separate memoryfor instruction and data.
19 Von Neumann and Harvard architectures Allows instructions and data to be mixed and stored in the same memory module;More flexible and easier to implement;Suitable for most of the general purpose processors (i.e. Pentium is using an von Neumann architecture for the external memory and is using a Harvard architecture for its internal cache memory (allowing 8KB of cache for data and anther (separate) 8kB of cache for instructions).Harvard:Uses separate memory modules for instructions and for data;It is easier to pipeline;Higher memory throughput;Suitable for DSP (Digital Signal Processors).
20 The Basic System Components Typical Von Neumann systemFor example, the 80x86 family uses the Von Neumann architecture) VNA).
21 The System Bus ModelA refinement of the von Neumann model, the system bus model has a CPU (ALU and control), memory, and an input/output unit.Communication among components is handled by a shared pathway called the system bus, which is made up of the data bus, the address bus, and the control bus. There is also a power bus, and some architectures may also have a separate I/O bus.
22 System busesA bus is a set of wires, that interconnects all the components (subsystems) of a computer.Source component sources out data onto the bus, the destination component inputs data from the bus.A system may have a hierarchy of buses.The I/O controller may use a second bus, often described as I/O bus or local bus to access a variety of attached devices.PCI bus is an example of a very common local bus.
23 Address busCPU reads/writes data from the memory by addressing a specific location; outputs the location of the data on the address bus; memory uses the address to access the proper data.Each I/O device (such as monitor, keypad, etc.) has a unique address as well (or a range of addresses); when accessing a I/O device, CPU places its address on the address bus. Each device will detect if it is its own address and act accordingly.Devices always receive data from the CPU; CPU never reads the address bus (it is never addressed).
24 Max Addressable Memory 80x86 Family Address Bus SizesIn WordsMax Addressable MemoryAddress Bus SizeProcessorOne Megabyte1,048,57620808880868018880186Sixteen Megabytes16,777,216248028680386sxFour Gigabytes4,294,976,2963280386dx8048680586 / Pentium (Pro)
25 Data busWhen the CPU reads data from memory, it first outputs the address on the address bus, then the memory outputs the data onto the data bus; the CPU reads the data from data bus.When writing data onto the memory, the CPU outputs first the address on the address bus, then outputs the data onto the output bus; memory then reads and stores the data at the proper location.The process to read/write to a I/O device is similar.
26 80x86 Processor Data Bus Sizes 808880188168086801868028680386sx3280386dx804866480586 class/ Pentium (Pro)
27 Control busAddress and data buses consist of n lines, which combine to transmit one n bit value; control bus is a collection of individual control signals.These signals indicate whether the data is to be read into or written out the CPU, whether the CPU is accessing memory or an IO device, and whether the I/O device or memory is ready for the data transfer.This bus is mostly a collection of unidirectional signals.