Computer Speed and Processor Evolution

Book chapter 2 Computer Organization and Architecture By William Stallings Pub.: Prentice-Hall International, Inc.

The evolution of computers has been characterized by increasing processor speed, decreasing component size, increasing memory size, and increasing I/O capacity and speed. Anything more??? HW.

One factor responsible for the great increase in processor speed is the shrinking size microprocessor components components; this reduces the distance between components and hence increases speed increases speed. However, the true gains in speed true gains in speed in recent years have come from the organization organization of the processor, including heavy use of pipelining and parallel execution techniques and the use of speculative execution techniques (tentative execution of future instructions that might be needed). shrinking size of All of these techniques are designed to keep the processor busy as much of the time as possible.

A critical issue in computer system design is Balancing the performance of the various elements so that gains in performance in one area are not handicapped by a lag in other areas. In particular, processor speed processor speed has increased more more rapidly than memory access time. A variety of techniques is used to compensate for this mismatch, including caches, wider data paths from memory to processor, and more intelligent memory chips.

Factors Affecting Processor Speed Factors Affecting Processor Speed There are many factors which affect how fast your computer can process data and instructions: The amount of RAM memory The speed and generation of your CPU (the system clock) The size of the Register on your CPU The Bus type and speed The amount of Cache memory Read & HW: http:// http://www.alf.sd83.bc.ca/courses/it12/using_it/processor_speed.htm www.alf.sd83.bc.ca/courses/it12/using_it/processor_speed.htm

How Registers Affect Speed How Registers Affect Speed A register is a high-speed memory area on the CPU, which hold data and instructions currently being processed. Most CPUs sold today, for both PCs and Macintosh computers, have 64-bit registers. The size of the registers size of the registers, which is sometimes called the word size word size, indicates the amount of data with which the computer can work at any given time.

The bigger the word size, the more quickly the computer can process a set of data. Occasionally, you will hear people refer to "32- bit processors," or "64-bit processors," or even "64-bit computers." This terminology refers to the size of the registers size of the registers in the processor. If all other factors are kept equal, a CPU with 32-bit registers can process data twice as fast as one with 16-bit registers.

Memory and Computing Power Memory and Computing Power The amount of RAM in a computer The amount of RAM in a computer can have a profound effect on the computer's power. More RAM means the computer can use bigger, more powerful programs, and those programs can access bigger data files. More RAM also can make the computer run faster. The computer does not necessarily have to load an entire program into memory to run it. However, the greater the amount of the program that fits into memory, the faster the program runs.

For example, a PC with 16 MB of RAM is able to run Microsoft Windows 98, even though the program actually occupies about 195 MB of disk storage space. When you run Windows, the program does not need to load all its files into memory to run properly. It loads only the most essential parts into memory.

When the computer needs access to other parts of the program on the disk, it can unload, or swap nonessential parts from RAM to the hard disk. Then the computer can load, or swap in code or data it needs. This process is called swapping While this is an effective method for managing a limited amount of memory, it can result in slow system performance because the CPU, memory, and disk are continuously occupied with the swapping process. If your PC has 64 MB of RAM (or more), you will notice a dramatic difference in how fast Microsoft Windows 98 runs because the CPU will need to swap program instructions between RAM and the hard disk much less often. swap out out, swap in, the program swapping. .

Computer's Internal Clock Computer's Internal Clock Every microcomputer has a system clock clock's primary purpose is not to keep the time of day. The clock is driven by a quartz crystal. When electricity is applied, the molecules in the crystal vibrate millions of times per second, a rate that never changes. The speed of the vibrations is determined by the thickness of the crystal. The computer uses the vibrations of the quartz in the system clock to time its processing operations. Over the years, system clocks have become steadily faster. E.g., the first PC operated at 4.77 megahertz. Hertz Hertz (Hz) (Hz) is a measure of cycles per second. Megahertz (MHz) "millions of cycles per second." system clock, but the Megahertz (MHz) means

The computer's operating speed is tied to the speed of the system clock. For example, if a computer's clock speed is 300 MHz, it "ticks" 300 million times per second. A clock cycle clock cycle is a single tick, or the time it takes to turn a transistor off and back on again. A processor can execute an instruction in a given number of clock cycles. As the system's clock speed increases, so does the number of instructions it can carry out each second. Clock speed has a tremendous impact on CPU performance. A CPU operating at 300 MHz can process data nearly twice as fast as the same one operating at 166 MHz. & increasing

Bus In microcomputers, the term bus path between the components of a computer. There are two main buses in a computer: the in internal (or system) bus the ex external (or expansion) bus The system system bus resides on the motherboard and connects the CPU to other devices CPU to other devices that reside on the motherboard motherboard. An expansion bus expansion bus connects external devices, the keyboard, mouse, modem, printer, etc., to the CPU Cables from disk drives and some other internal devices may also be plugged into the bus. bus refers to the external devices, such as to the CPU.

The system bus system bus has two parts: the data bus the address address bus bus The address bus address bus leads from the CPU to RAM. The data bus data bus connects the CPU to memory well as all the storage storage, input/output, and communication devices. data bus and CPU to RAM. CPU to memory as

..Bus ..Bus: : System Bus System Bus Data Data Bus Bus The data bus i data bus is an electrical path that connects the CPU, memory, and the other hardware CPU, memory, and the other hardware devices devices on the motherboard. Actually, the bus is a group of parallel wires. The number of wires in the bus number of wires in the bus affects the speed at which data can travel between hardware components, just as the number of lanes on a highway affects how long it takes people to reach their destinations.

Because each wire can transfer 1 bit of data each wire can transfer 1 bit of data at a time, an 8-wire bus can move 8 bits at a time, which is a full byte. A 16-bit bus can transfer 2B, and a 32-bit bus can transfer 4B at a time. Newer model computers have a 64-bit data bus, which transfers 8 bytes at a time. With a wider bus, the computer can move more data in the same amount of time (or the same amount of data in less time).

..Bus: System Bus Address Bus The address bus address bus is a set of wires similar to the data bus. The address bus connects only the CPU and RAM CPU and RAM and carries only memory addresses addresses. Remember, each byte in RAM is associated with a number, which is its memory address. Q. In theory, today's CPUs have address buses that are wide enough to address 64 GB of RAM. arries only memory

Requests for data are sent from the CPU to RAM along the address bus. The request consists of a memory address. The data comes back to the CPU via the data bus.

..Bus Types PC buses PC buses are designed to match the capabilities of the devices attached to them. When CPUs could send and receive only 1 byte of data at a time, there was no point in connecting them to a bus that could move more data. As microprocessor technology improved, however, chips were built that could send and receive more data at once, and improved bus designs created wider paths through which the data could flow.

Common bus technologies include: Industry Industry Standard Architecture (ISA) bus Standard Architecture (ISA) bus Local bus Peripheral Component Interconnect (PCI) Peripheral Component Interconnect (PCI) bus Accelerated Graphics Port (AGP) Accelerated Graphics Port (AGP) bus Universal Serial Bus (USB) Universal Serial Bus (USB) IEEE IEEE 1394 1394 bus bus

The Industry Standard Architecture (ISA) bus Industry Standard Architecture (ISA) bus is a 16-bit data bus. ISA was the industry standard on its release in the mid-1980s and is still used in many computers to attach slower devices (such as modems and input devices) to the CPU. ISA is a type of bus that carries information between an ISA style expansion slot and the CPU.

One 8-bit and five 16-bit ISA slots on a motherboard Variants PC/104 - Embedded variant of ISA Low Pin Count (LPC) - Low pin count version of ISA Extended Industry Standard Architecture (EISA) Micro Channel architecture (MCA) Wiki

Local bus Local bus technology was developed to attach faster devices to the CPU. A local bus is an internal system bus an internal system bus that runs between components on the motherboard. Most system buses use some type of local bus technology today and are coupled with one or more kinds of expansion bus.

Peripheral Peripheral Component Interconnect (PCI) Component Interconnect (PCI) bus bus is a type of local bus designed by Intel to make it easier to integrate new data types, such as audio, video, and graphics. Most new computers use a PCI bus and PCI expansion slots. A PCI bus connects PCI expansions cards to the CPU. Since June 1992.

Visit: Visit: http://pcisig.com/ - the community responsible for developing and maintaining the standardized approach to peripheral component I/O data transfers. 133 MB/s (32-bit at 33 MHz the standard configuration) 266 MB/s (32-bit at 66 MHz or 64-bit at 33 MHz) 533 MB/s (64-bit at 66 MHz) Three 5-volt 32-bit PCI expansion slots on a motherboard (PC bracket on left side)

Accelerated Graphics Port (AGP) Accelerated Graphics Port (AGP) bus bus incorporates a special architecture that allows video cards video cards to access the system's RAM directly directly, greatly increasing the speed of graphics performance. The AGP standard has led to the development of many types of accelerated video cards that support 3-D and full-motion video. While AGP improves graphics performance, it cannot be used with all PCs. The system must use a chip set that supports the AGP standard. Most new computers feature AGP graphics capabilities in addition to a PCI system bus and an expansion bus. system's RAM

Universal AGP slot (brown, top) and PCI 2.2 slot (white beige, bottom)

Q. What is the difference between AGP and PCI? The biggest difference between AGP and PCI PCI graphics cards is that AGP AGP cards can access the system memory to help with complex operations such as texture mapping. PCI PCI cards can only access the memory available on the actual card. AGP AGP doesn't share bandwidth with other devices, whereas PCI PCI cards do. AGP and HW: https://pc.net/helpcenter/answers/difference_between_agp_and_pci

Universal Serial Bus (USB) Universal Serial Bus (USB) - not only provides fast data transfer speeds, it also eliminates the need for expansion slots and boards. Most new PCs and Macintosh computers feature at least one USB port, and each USB port can support 127 different devices. If you have USB-compliant devices such as key boards, mice, printers, and modems, you can plug them all into a single USB port.

Length: 25 m (6 ft 7 in16 ft 5 in) (by category) Width: 12 mm (type-A) | 8.45 mm (type-B) | 6.8 mm (mini/micro) | 8.25 mm (type-C) Height: 4.5 mm (type-A) | 7.26 mm (type-B) | 10.44 mm (type-B SuperSpeed) | 1.8 3 mm (mini/micro) | 2.4 mm (type-C)

Different USB connectors: From left to right: (1) male Micro USB B-Type, (2) proprietary UC-E6, (3) male Mini USB (5-pin) B-type, (4) female A-type, (5) male A-type, (6) male B-type. Shown with a centimeter ruler. Female A-type connector (4th from left) is "upside down" to show the pins. See: https://en.wikipedia.org/wiki/USB

HW & Read Micro USB Live USB Wireless USB USB flash drive

IEEE 1394 IEEE 1394 IEEE 1394 IEEE 1394 is an interface standard for a serial bus for high-speed communications and isochronous real-time data transfer. It was developed in the late 1980s and early 1990s by Apple, which called it FireWire The 1394 1394 interface is also known by the brand i.LINK (Sony), and Lynx (Texas Instruments). FireWire. 9-pin FireWire 800 connector Alternative Ethernet-style cabling used by 1394c

Traditionally, the performance of computer performance of computer buses was measured by the number of bits they buses was measured by the number of bits they could transfer at one time. could transfer at one time. Hence, the newest 64-bit buses are typically considered the fastest available. However, buses are now also being measured according to their data transfer according to their data transfer rates - - the amount of data they can transfer in a second. This type of performance is usually measured in megabits per second (Mbps) megabits per second (Mbps) or megabytes per second (MBps). rates

For example, A USB bus has a data transfer rate of 12 Mbps. An IEEE 1394 bus has a data transfer rate of 400 Mbps. AGP buses are typically rated at 266 MBps but can support data transfer rates of more than 1 GBps. PCI buses offer data transfer rates of 133 MBps. Some manufacturers also rate the speed of their also rate the speed of their system buses in megahertz. system buses in megahertz. For years, system buses ran at a speed of 66 MHz; contemporary systems offer bus speeds of 100 MHz and 200 MHz. When coupled with a fast processor, a high-speed bus can result in an exceptionally high- performance system.

Cache Memory Moving data between RAM and the CPU's Moving data between RAM and the CPU's registers registers is one of the most time operations a CPU must perform, simply because RAM is much slower than the CPU. A partial solution to this problem is to include cache memory cache memory. Cache (pronounced cash) memory is extremely fast memory, which hold the most recently used data and instructions. Why so fast? time- -consuming consuming

When a program is running and the CPU needs to read data or program instructions from RAM, the CPU checks first to see whether the data is in cache memory. If the data is not there, the CPU reads the data from RAM into its registers, but it also loads a copy of the data into cache memory copy of the data into cache memory. The next time the CPU needs that same data, it finds it in the cache memory and saves the time needed to load the data from RAM. but it also loads a

There are 2 types of cache memory, L1 and L2 The difference being their speed and size. L1 L1, or level 1 cache is much faster holds a lot less data less data than L2, or level 2 cache. Since the late 1980s, L1 cache memory has been built into the CPU. The first CPU caches came with 0.5 KB, then 8 KB, then 16 KB 32 KB 256 KB 512 KB L1 cache built in. faster, but also

L2 cache can either be "on die", meaning it is built into the CPU, or it can be built into the motherboard. The closer closer cache memory is to the CPU, the faster faster it is. Therefore, a CPU with both L1 and L2 cache on die will be faster than a CPU with its L2 cache on the motherboard. Many PCs sold today have 512 KB, 1024 KB, or 2 Mb of L2 cache memory.

The cache speeds up processing by storing frequently used data or instructions in its high- speed memory. External (Level-2) cache is shown here, but most computers also have internal (Level-1) cache memory circuitry built into the CPU. Read: http://www.hardwaresecrets.com/how-the-cache- memory-works/3/

HW: PPT 10min https://es.cs.uni- kl.de/publications/data/Jede15.pdf Eric Jedermann, Exposed Datapath Processor Architecture Implementation Survey

Harvard vs. Princeton / Von Neumann architecture?

Institute for Advanced Study (IAS IAS) The IAS electronic computer for Advanced Study (IAS Jersey, U.S. It is sometimes called the von Neumann machine, since the paper describing its design was edited by John von Neumann, a mathematics professor at both Princeton University and IAS IAS. IAS machine was the first computer to be built at the Institute IAS) in Princeton, New wiki

Q. Mention 5 things you need to consider for this simple architecture. What was in the mind of Von Neumann & Co. while designing this?

1. CA Because the device is primarily a computer, it will have to perform the elementary operations of arithmetic most frequently arithmetic most frequently. + - x / It is therefore reasonable that it should contain specialized organs for just these operations. It must be observed, however, that while this principle as such is probably sound, the specific way in which it is realized requires close scrutiny. At any rate a central arithmetical part of the device will probably have to exist and this constitutes the first specific part: CA.

2. CC The logical control logical control of the device, that is, the proper sequencing sequencing of its operations, can be most efficiently carried out by a central control organ.

3. M Any device which is to carry out long and complicated sequences of operations (specifically of calculations) must have a considerable memory memory . . . The instructions instructions which govern a complicated problem may constitute considerable material, particularly so, if the code is circumstantial (which it is in most arrangements). This material must be remembered. At any rate, the total memory total memory constitutes the third specific part of the device: M.