Higher Computing: COMPUTER SYSTEMS Part 2: Computer Structure — 6 hours © Marr College Higher Computing Slide 1Five box diagram Computer Structure The five box diagram represents the basic components of a computer system. It represents input devices, processor, main memory, output devices and backing storage. Processor Main Memory YT] Backing Store Marr College Higher Computing Slide 2Computer Structure The purpose of the processor The processor - also known as the central processing unit (CPU) - is the ‘brain’ or ‘engine’ of the computer system. Its purpose is to interpret instructions and process data contained in computer programs. Marr College Higher Computing Slide 3Computer Structure Parts of a processor The CPU consists of three main parts: The control unit has a timer to that sends signals to fetch, decode and execute program instructions. The arithmetic and logic unit performs arithmetic calculations e.g. / * + - and logic operations e.g. AND, OR etc. The registers are temporary storage areas that hold data being processed; instructions being executed; and addresses to be accessed. Marr College Higher Computing Slide 4Computer Structure Ea) The purpose of the ALU The arithmetic and logic unit is a digital circuit that performs arithmetic and logical operations. The ALU can perform many operations including: Integer arithmetic e.g. addition and subtraction Bitwise logic operations e.g. AND, OR, NOT, XOR * The inputs to the ALU are the data to be operated on (called operands) * and a code from the control unit indicating which operation to perform. Operation (opcode) Input (operand) = Its output is the result of the ‘omputation. Output (result) semputeson Marr College Higher Computing Slide 5Computer Structure The purpose of the Control Unit The control unit is the circuitry that controls the flow of data through the processor, and coordinates the activities of the other units within it. It can be described as the "brain within the brain", as it controls what happens inside the processor, which in turn controls the rest of the PC. CONTROL UNIT ¥ It performs the tasks of fetching, decoding, managing execution and then storing results and has a timer to synchronise these events. Marr College Higher Computing Slide 6Computer Structure The Address Bus * Carries address information from the CPU to main memory and any other attached devices It is uni-directional i.e. one-way only * The width (i.e. the number of wires) determines the number of memory locations the CPU can address A 32 bit address bus has 32 parallel wires each switched on (1) or off (0) that can address locations starting from: 0000 0000 0000 0000 0000 0000 0000 0000 (decimal 0) up to and including address: 1114 1111 1111 1111 1111 1111 1111 1111 (decimal 232 -1) making a total of 2? addresses Every time a wire is added to the width of the address bus, the address range Marr College Higher Computing Slide 7Computer Structure The Data Bus * Carries data to and from the CPU, main memory and any other devices attached * It is bi-directional i.e. two-way * The number of wires determines the quantity of data that the bus can carry so increasing the number of wires in the data bus increases the quantity of data it can carry * Atypical 32-bit data bus can carry 32-bits of data or instructions at a time In Higher Computing we make the assumption that: The size (number of wires) on the data bus determines the size of the memory locations in RAM e.g. A 32 bit data bus means each memory location in RAM stores 32 bits (i.e. 4 bytes equivalent) Marr College Higher Computing Slide 8Computer Structure The Control Bus - control lines & Signals are sent out and received on the control bus. The control bus is not really a ‘bus’ as it does not transfer data or addresses. It is made up of discrete wires each with a specific function: Read and Write signals are initiated to fetch - execute instructions in memory CPU in step (clock rate measured megahertz or gigahertz) cleared and the machine reboots. Interrupts are signals usually from I/O devices that halt program execution temporarily. The CPU may ignore them e.g. printer out of paper Non-maskable Interrupts cannot be ignored e.g. power failure Marr College Higher Computing Clock line carries a series of clock pulses at a constant rate to keep the Reset halts the execution of the stored program. Internal registers are Slide 9The buses: Address, Data and Control Address bus Memory HE Adress Other Register registers a 2 re] Data bus = Memory Data o Register 3 = + 5 < || Control Bus (Read / Wri 5 Aritimetle ‘ontrol Bus (Read / Write) g i 3 and Logic Control Unit a Unit (ALU) = E ° ° Clock pulses Electronic clock Marr College Higher Computing Slide 10Computer Structure The fetch-execute cycle & The fetch-execute cycle of the processor refers to the sequence that is completed for each instruction in a program. Fetch Sequence Move the value in the program counter to the memory address register Send the value in the memory address register to memory via the address bus Return the value stored in memory via the data bus Store the value in the memory data register Copy the instruction from the memory address register to the instruction register Increment the program counter. The instruction in the instruction register is then Decoded Execute Sequence Instruction is executed Marr College Higher Computing Slide 11Computer Structure The purpose of the Registers The registers are small, fast storage areas that temporarily hold data. Instructions or addresses, The registers in the CPU have three main functions. They are to hold data being processed, instructions being executed, and addresses being accessed. memory address register (MAR) - holds the address of a location in memory memory data register (MDR) - holds data just read from or written to memory program counter (PC) - holds the address of the next instruction to be fetched Instruction register (IR) - holds the current instruction being executed ras general purpose registers - can be used by programmers Marr College Higher Computing Slide 12Computer Memory - Cache Cache Memory Is a small area of ‘super fast’ access memory, between the processor and main memory, which stores frequently used instructions and data. 2 1. Cache fetches data from next to 3 CPL’ checks to see whether 2 current addresses in main memory ext (nstuction I requires oO is in cache 3 z 5 6 - 5 3 2 cpu E ° 6 4. If not, the CPU has to fetch next 3. If it is, then the instruction is instruction from main memory - a fetched from the cache - a very much slower process fast position Marr College Higher Computing Slide 13Main memory - RAM Main memory is commonly referred to as RAM (random access memory). RAM is used to hold program instructions and data before and after processing by the CPU. RAM is volatile i.e. loses it contents when switched off. Reading from RAM is slower than accessing registers or cache. Use of cache avoids slower fetches from Computer Structure RAM chip Increasing memory size (capacity in Mb/Gb) improves system performance as more programs and data can be held. Slide 14 Marr College Higher ComputingComputer Structure Main memory - ROM Another type of memory is Read Only Memory (ROM). ROM is used to store the bootstrap loader program that locates the operating system on the hard disc when the computer ‘boots up’. Features - ROM * ROM data is permanently etched on chip * Read-only so data cannot be changed * Data not lost when computer switched off ROM chip Marr College Higher Computing Slide 15Backing storage Backing storage is where the computer permanently saves computer programs and data. Computer Structure This includes devices such as the hard disc and media such as DVD. Slide 16 Marr College Higher ComputingDistinguishing between different parts of memory Registers - fastest CPU Cache - slower than registers but fast as no READ needed Bandwidth ‘ LE , Latency & Size Inerases Increase Tots feta RAM - slower than registers and cache System DRAM Computer Structure Backing store - slowest speed of access Lbs el dal gel] (hard disk, tape, etc.) Marr College Higher Computing Slide 17Memory Computer Structure Investigate the following elements of computer memory: Registers cache main memory backing storage Distinguish between the above elements of memory according to: Function speed of access Produce a brief word processed report or powerpoint presentation. Include an image/diagram of each component. Marr College Higher Computing Slide 18Addressability Main memory consists of a number of storage locations, each of which is F Agoress M1 identified by a unique address. 0100000i- The ability of the CPU to identify each | location is known as its | Acres 12 addressability. 0i0000i0- Each location stores a word i.e. the Aggress M3, number of bits that can be processed | by the CPU in a single operation. Word 010010 length may be typically 16, 24, 32 or | as many as 64 bits. | Acoress mt Computer Structure A large word length improves system performance, though may be less efficient on occasions when the full word length is not used Marr College Higher Computing Slide 19Computer Structure Calculating memory capacity of a computer Memory capacity can be calculated if we know: * the number of lines on the address bus * and the number of bits stored in each memory location Note: in Higher we assume the number of bits in a memory location to be the same as the number of bits the data bus can carry. Formula Amount of storage locations = 2 width of the address bus Memory capacity = 2 the width of tne address bus * width of data bus Example A computer has a 24 bit address bus and a 16 bit data bus. Calculate the maximum amount of memory this computer can use. 224 * 2 bytes = 33,554,432 bytes 33,554,432 / 1024 = 32,768 kilobytes 32768 / 1024 = 32 megabytes Marr College Higher Computing Slide 20Calculating memory capacity of a computer Gaza Calculate the total memory requirements of the following computer systems: a) Processor has a 16-line address bus and each location stores 16 bits. b) Processor has a 16-line address bus and each location stores 32 bits. c) Processor has a 32-line address bus and a 24-line data bus. d) Processor has a 36-bit address bus and a word length of 32 bits. Computer Structure Remember: Location size = data bus size = word length / size © Marr College Higher Computing Slide 211 Desktop computer Processing Power - measured by clock speed e.g, 3 Ghz. Memory size - typically 3 Gb. Backing storage - large hard disc typically 750 Gb - 1 Tb. Also, CD-RW and DVD-RW. Input devices - keyboard, microphone, mouse, webcam etc. Output devices - monitors, j printers, speakers, modem etc Computer Structure Typical uses - applications (word, excel), email, internet, gaming Marr College Higher Computing Slide 22Computer Structure Laptop computer Marr College Processing Power - measured by | clock speed e.g. 1.75 Ghz - 2.4 GHz Memory size - typically 512 Mb — 1 Gb. Backing storage - large hard disc typically 40 - 100 Gb. Also, CD- RW and DVD-RW. Input devices - keyboard, touchpad Output devices - LCD integrated monitor, printers, speakers, modem etc Typical uses - applications (word, excel), email, internet, gaming. Portable to work on the move. Higher Computing Slide 23Palmtop computer Computer Structure Processing Power - measured by clock speed e.g. 200 - 500 Mhz Memory size - typically 32 - 64 Mb, Can extend with memory card. Backing storage - Same as memory (battery powered) Input devices - stylus and touch screen Output devices - integrated LCD screen Typical uses - personal organiser and connectivity to PC. Marr College Higher Computing Slide 24Mainframe computer A large and powerful computer that deals with very high volumes of data processing. Features Processing power: several processors RAM: 32 GB or more Backing storage: 100s of GB, tape drives I/O: keyboard, printers, monitors Computer Structure Uses Used by large organisations to provide remote access to central computer via terminals e.g. a bank’s ‘cashpoint’ machines are directly connected to a mainframe at head office. © Marr College Higher Computing Slide 25Computer Structure 3) Embedded computer Embedded computers are special-purpose systems where the computer is embedded within the machine it controls e.g. a computer system in a car. They perform a specific task Considered a Marr College Higher Computing Slide 26Computer Structure System Performance: Clock Speed Clock Speed Is the number of clock pulses a CPU generates per second. Measured in gigahertz i.e. 1GHz = 1 billion pulses per second. These pulses synchronise the steps of the fetch-execute cycle e.g. a clock pulse starts a ‘fetch’, or Q triggers placing data in the MDR. The faster the clock speed - the more operations can be executed per second. Clock speed indicates processing power but some instructions need more clock pulses on one processor than on another. Marr College Higher Computing Slide 27Computer Structure System Performance: MIPS rate & Millions of Instructions per Second (MIPS) MIPS roughly measures the number of machine code instructions that a CPU can execute per second. Pros Rough measure of processor performance e.g. Intel 8080: 500 KIPS at 2 MHz Intel 486 DX2: 54 MIPS at 66 MHz Intel Pentium III 1354 MIPS at 500 MHz Intel Core 2 X6800: 27,079 MIPS at 2.93 GHz Cons No standard way of measuring MIPS as does not take into account complexity of an nstruction e.g. some instructions require more time than others. Refers only to CPU speed, not other factors such as I/O Therefore a machine with a high MIP rate may not run an application faster than a CPU with a lower rate due to other limiting factors. Marr College Higher Computing Slide 28Computer Structure System Performance: MIPS rate & How to calculate a CPU’s MIPS rating There is a simple formula to calculate MIPS. Formula MIPS = n/ (t x 1,000,000) Where n is the number of instructions executed and t Example No standard way of measuring MIPS as does not take into account complexity of an nstruction e.g. some instructions require more time than others. Refers only to CPU speed, not other factors such as I/O Therefore a machine with a high MIP rate may not run an application faster than a CPU with a lower rate due to other limiting factors. Marr College Higher Computing Slide 29System Performance: MIPS rate Millions of Instructions Per Second (MIP) A clock speed of 200 MHz does not mean that 200 million instructions are executed per second. Because of this, clock speed isn’t an accurate measure of performance. It may therefore be the case that two processors have the same clock speed but different MIP rates. In order to get a better picture of performance we measure the CPUs ability in how many million instructions it can process per second, also referred to as the machine cycle time. Computer Structure Marr College Higher Computing Slide 30System Performance: MIPS rate Working out MIPS rate It can take at least five clock pulses to execute an instruction. Example — a CPU with a clock speed of 200 MHz => > 200,000,000 clock pulses / 5 pulses per instruction ==> 40,000,000 instructions per sec. Thus CPU = 40 MIPS Computer Structure Marr College Higher Computing Slide 31System Performance: MIPS rate Problems with MIPS rate + Using MIPS rate as a comparison factor also has problems. + MIPS rate depends on what sort of instructions are being carried out + There is no standard set and so some manufacturers could use instructions which require less clock pulses to complete thus giving the impression of a faster MIPS rate than it actually is. Computer Structure Marr College Higher Computing Slide 32 3)Computer Structure System Performance: FLOPS Floating Point Operations per Second (FLOPS) Measures how many floating point operations are processor can perform per second. a FLOPS are more accurate than MIPS. They measure a number of clearly definable arithmetic tasks carried out per second. Marr College Higher Computing Slide 33System Performance: Application-based Test (Benchmarks) Benchmarks This involves running various application programs on different computer systems and observing which system runs these programs the fastest. Most reliable measure of processor performance because it provides the user with actual evidence of how well different processors perform complex operations at high speed. Computer Structure cial il Marr College Higher Computing Slide 34Factors that affect system performance Data bus width - Increasing the data bus width increases processor throughput i.e. an 8-bit bus can transfer 1 byte of data; a 16-bit can transfer 2 bytes etc. Cache memory - Using / increasing cache (SRAM) means less ‘fetches’ from slower main memory (DRAM). Peripheral speed - Slow transfer rates slow the system down so better to choose peripherals with faster transfer rates. Computer Structure Marr College Higher Computing Slide 35Current trends in computer hardware Faster clock speeds - improved CPU throughput as millions more instructions per second, More RAM - more sophisticated software can be run, and improved multitasking i.e. more programs loaded simultaneously. More backing storage capacity means more data can be stored permanently. Computer Structure © Marr College Higher Computing Slide 36