LAST MAJOR UPDATE: 10/4/04; 9/30/06; 10/25/07
Constantly Under  Construction!
I am developing a MAJOR REWRITE of this,
with better graphics, to make it more suitable for independent learning, and more efficiently Web oriented.
Note that the blinking text designates things that I need to work on; the material is not wrong, but can be improved.
(Don't worry, I don't like blinking text, either, so there will not be any in the finished product!)

LEARNING MODULE IIIC
PERSONAL COMPUTERS

        Personal computers are stand-alone, general purpose, microprocessor-based computers. They are subdivided into portables (including laptops, hand-held devices, etc.),  and desktops.   Originally microcomputers could support only one user; however, the newer generations of microprocessors can be accessed, simultaneously by several users. Therefore, it is important to distinguish between "personal computers" (PCs") like those based on Intel/AMD/Cyrix or Motorola microprocessors which the user normally have all to himself, and "multiuser computers" ("MCs") which are, by design, a single processor or multiprocessor complex shared by numerous dumb terminals. (It should be noted that modern micros are capable of supporting multiple users, so it is how they are used, usually indicated by the type of Operating System employed, that makes them either PCs or MCs) This LM is separated from LM IIIB in order to distinguish concepts primarily to microcomputers from those that apply to generic computers (covered in LM IIIB), so our goals of this Learning Module IIIC are to help the student:

1. learn the basic architecture of microprocessor based personal computers.
2. learn the distinction between different kinds of microprocessors.
3. understand the symbiotic relationship between the CPU and the software designed for it.
4. understand the advantages of platform portability.
5. appreciate the advantages of distributed computing.

TPQ 1: Rewrite the preceding objectives in terms of personal accomplishments to be attained after finishing the study of this learning module.

The sequence of presentations in this learning module is as follows.  You can click on any link to jump directly to that section.

1. MICROCOMPUTER COMPONENTS
2. MICROPROCESSOR AND CHIPSET ARCHITECTURES
3. MICROCOMPUTER BUSES
4. THE HARDWARE-SOFTWARE DILEMMA
5. TOTAL PORTABILITY, THE PERFECT WORLD OF THE FUTURE
6. DISTRIBUTED COMPUTING, THE FUTURE OF COMPUTING
7. SUMMARY  {Convert to FIB?}

A good reference for this whole section is:
Building a PC by Morris Rosenthal
Watch the excellent (though oversimplified) brief,  "Tour of a PC" from How Stuff Works.

1. The System Unit case (See Figure PC-1B, noting that cases come in three basic sizes.) houses the
1. Motherboard (main circuit board) is a printed circuit board where the electronic circuits and buses, printed on both surfaces of the board itself, interconnect the motherboard components.  (For descriptions and pictures of current motherboards see Our Best Selling Motherboards from TigerDirect.com.)
   
1. PC motherboards have three basic layouts (called "form factors"): AT, ATX, and NLX.   These specify the size and shape of cases and motherboards which must to match, i.e. if you have an ATX case you must install an ATX motherboard and power connectors on the case have to match  the motherboard.
1. AT, the original form factor, was created by IBM for its PC AT, a 286-based machine. The AT layout was superceded by the "baby AT".
2. ATX, smaller than the AT motherboard, improved on its back plane circuitry and component interfaces making it more efficient, e.g. the disk drive cable connectors are nearer to the drive bays and the CPU is closer to the power supply and cooling fan.  It has more space for expansion slots and built in ports reduce the number of ribbon connectors needed.  The Micro ATX is a smaller version of the ATX layout with fewer expansion slots.
3. NLX, a factor from the late 90's, is designed to accommodate newer components and to make it easier to upgrade your computer.
4. BTX (from From AT to BTX: Motherboard Form Factors), or Balanced Technology Extended form factor, unlike its predecessors is not an evolution of a previous form factor but a total break away from the popular and dominating ATX form factor. BTX was developed to take advantage of technologies such as Serial ATA, USB 2.0, and PCI Express. Changes to the layout with the BTX form factor include better component placement for back panel I/O controllers and it is smaller than microATX systems. The BTX form factor provides the industry push to tower size systems with an increased number of system slots.
1. For a excellent overview of the evolution of PC form factors see: From AT to BTX: Motherboard Form Factors.
   
2. From a simplified viewpoint, a typical motherboard contains the following components:
1. mounts for one or more microprocessors, and the associated core-logic chipset,
2. mounts for primary memory (_________ and _________(1)) chips which may be directly plugged into the motherboard.  RAM is typically packaged in SIMM or DIMM modules which plug into slots in the motherboard (See section 1.A.),
3. expansion slots in which expansion cards (see below) may be plugged to add devices to the system,
4. interfaces to peripheral devices and I/O ports,
5. clock/timer chips,
6. numerous other support chips, and
7. ________(2) which interconnect all the motherboard components. (See section 3, below.)
3. Examples:
1. See Figure PC-2 for a modern motherboard and its ports for peripheral devices. 
2.   Figure PC-2B shows a dual processor motherboard that can perform SMP (symetric multiprocessing) with Linux, BeOS, AND Windows NT, 2000, and XP; compare this to the dual overclocked Celeron motherboard used in the video shown in class.)    (links with strikethroughs are outdated or (worse) now unavailable (UGGGGG!); I am looking for alternatives.  If you find any, let me know on the class forum!)
3. For "state of the art" hardware, check out the latest "ultimate gaming machine" at:
1st Place Gaming Computer: CGW Editors Choice Gaming Computer (1st place in every gaming benchmark)
4. To get the latest on motherboard technology, explore the pages of
4. ABIT at: http://www.abit-usa.com/
5. ASUS at: http://usa.asus.com/index.aspx
2. Daughterboards is a generic term for circuit boards which plug into special slots on the motherboard to allow addition of upgraded CPUs or accelerator cards which contain coprocessors that speed up processing.
3. Expansion Cards plug into the expansion slots allowing modular expansion. The slots are normally connected to the CPU via buses. The expansion cards can provide system extensions such as:
1. peripheral devices.  See section 1.1.B, below.
2. coprocessors such as graphics accelerators that can execute specialized instructions releaving the CPU of this responsibility.  Separate graphics processors are becomming so prevalent that a new term "GPU" is used for a "graphics PU", execplified by  the processors built by NVIDIA and ATI.  Such GPUs exploit parallel processing and VRAM (video RAM) to draw and render 2D and 3D images.  GPU is functionally a second (or third) CPU, so it needs to share privileged access to main memory with the CPU(s).
3. additional RAM (______ or ______(3) modules)  Review LM IIIB, section 2.B.
4. networking.
4. Power Supply;
5. Internal secondary storage (See Learning Module V, SECONDARY STORAGE) devices that are inserted into bays (spaces reserved within the system case for internal devices that, when mounted, can be accessed from the front of the PC.):
1. Internal hard disks (gigabyte range) are the primary devices currently used to store software and data.
2. Internal Removable Disk Drive(s):Direct access storage is currently dominated by magnetic media (hard disks, removable hard disks, and floppies), but magneto-optical and read/write optical media (CD-RW, DVD, DVD-RAM, and DVD+RW) promise to revolutionize storage technologies!
6. Internal  CD-ROM drives are now common, especially on "multimedia PCs".  Note that these are more primitive than the optical media mentioned in section e.ii, above, but since CD-ROM drives are read only they are not included with the read/write secondary storage devices above.  However, since secondary storage optical drives can read CD-ROMs, only one optical drive is really necessary.
   
2. Peripheral Devices, which may be added to the system, are covered in subsequent LMs; they fall into four general categories.
1. Input devices (See LM V Input & Output.)including keyboards, pointing devices, etc.
2. Output devices (See LM V Input & Output.)including monitors, audio speakers, printers, etc.
3. Secondary storage Devices (See LM IV Secondary Storage)including external hard disks, removable storage drives, CDs, etc.
4. Communication devices (See LM VI Comm. & Networking.)include external modems, network cards, etc.
3. For an informative article on motherboard components see Tour Around the Motherboard from Building a PC by Morris Rosenthal.

1.2 Architecture and Organization of a typical Micro System:

1. A schematic of a generic microcomputer is given in Figure PC-3 which represents it as a collection of components that communicate via computer buses.
2. In general, a computer bus is a set of conductors that provides a common pathway between multiple electronic chips.
1. We learned in Learning Module IIIB, section 3.3, that buses carry three types of signals (_____, __________, and _________(4)), but, in fact, all of these are integrated within a single bus.
2. Also, the oversimplified layout of Figure PC-3 shows only one idealized bus which links the different motherboard components, but it actually contains several different types of buses.
3. Microcomputer buses are discussed in section 3, below, after the microprocessor properties, necessary to understand bus architecture, are presented.
4. The bandwidth of a bus is the number of megabytes it can transmit per second.  It can be calculated by multiplying the bus "width" (i.e. data capacity measured in bits) times its speed (measured inmegahertz), e.g. the PCI bus is 32 bits wide and runs at a speed of 33MHz so its bandwidth is _________(5).
3. In modern microcomputers the microprocessor and primary memory actually communicate with the motherboard buses via the core-logic chipset.  This chipset serves as a switching mechanism that dirrects motherboard bus traffic among the different devices that make up the system.
1. A chipset is an integrated set of complementary computer chips that are designed for a specific task.  In the case of the core-logic chipset, they coordinate the communication between the microprocessor, primary memory, and the various motherboard _________(6).
2. (from Wikipedia) The term "chipset" often refers to the two main motherboard chips, northbridge and southbridge. . The manufacturer of a chipset often is independent from the manufacturer of the motherboard. Examples of manufactors of PC motherboard chipsets include NVIDIA, ATI, VIA Technologies, SiS and Intel.
1. The northbridge (also known as the Memory Controller Hub or MCH) typically handles communications between the CPU, RAM, AGP or PCI Express, and the southbridge.
2. The southbridge, (also known as the I/O Controller Hub or ICH) is a chip that implements the "slower" capabilities of the motherboard. The southbridge can usually be distinguished from the northbridge by not being directly connected to the CPU. Rather, the northbridge ties the southbridge to the CPU.  (See Wikipedia's illustrations here. and motherboard schematic, below and the illustrations in Overview of PCIe from ARS Technica.)

NORTHBRIDGE (MCH) AND SOUTBRIDGE (ICH) BUS SCHEMATICS

Generic PCI Schematic	PCI Schematic for Intel Processors	Equivalent Schematic for the Mac Pro
(See Role of the Microprocesor Schematic, below.)	view larger image
From Overview of PCIe from ARS Technica.	From Intel® 850E chipset.	From Apples Mac Pro Website

3. A particular core-logic chipset is usually referred to by its proper name, e.g. the Intel® 850E chipset (See the enlarged schematic, above middle.)
4. For more information:
1. Wikipedia's links and references on it's chipset page.
   
2. Intel chipsets see http://www.intel.com/design/chipsets/index.htm
1. Peripheral devices are connected to the system unit by connecting them to external busses (either serial or parallel) that plug into computer ports, typically on the back of the system case.   See Figures 4-28 and 4-29 in the text.
1. The ports are connected to the motherboard via hardware interfaces that are either serial or parallel depending on the type of peripheral device being connected.  Note: the "disk controller" in Figure PC-3 is also an interface.
1. Serial ports and serial busses transmit data one bit at a time.   See Figure 4-30 in the text.
2. Parallel ports and busses transmit data several bits at a time, typically eight bits simultaneously over parallel wires. See Figure 4-31 in the text.  Parallel ports are being replaced by the new high speed USB and FireWire ports; see the discussion of these under external busses, to which the ports connect.
   
2. Some of these interfaces are sockets connected directly into the motherboard (e.g. serial, parallel, built-in USB, etc.) while others are wired to expansion cards that are inserted into the _________(7).
3. Note that USB 2.0 and FireWire ports are replacing traditional ports (serial, parallel, SCSI, audio, keyboard, mouse, etc.)  See the nice illustration of this in Figure 4-32 in the text.
4. Special Purpose Ports (not normally provided on standard PCs) include:
1. MIDI (Musical Instrument Digital Interface) ports allow the input and output to any musical instrument capable of electrical I/O.
2. IrDa (Infared...) ports allow the wireless input and output via infared; IR transmission require unubstructed line of site.
3. Bluetooth ports allow wireless input and output via radio waves; this technology does NOT require line of site. so it is much more versitle than IrDa connections.  Bluetooth is designed to provide wireless communication between independent devices, particularly portable devices like notebook computers, PDAs, cell phones, etc.  See LM VI for more details.
4. SCSI (pronounced "skuzzy") ports provide parallel connections to external devices; this technology is being replace by FireWire; see the discussion of these under external busses, to which these ports connect.

2. MICROPROCESSOR ARCHITECTURE:

1. MPs, differing greatly in architecture, can be loosely classified according to their speed.  In additon to the actual design of the chip, the speed of a MP (or CPU) depends on at least four things (the larger the faster)
1. word size (the number of bits in the CPU _________(8))
2. data bus size (number of data bits simultaneously transferable)
3. address bus size  (number of address bits , i.e. the size of addressable memory), and
4. clock speed (measures in megahertz or 1000 "ticks" per second).  "Overclocking" is a new technique, popular with hackers and action game fanatics, that allows microprocessors to be run at higher clock speeds than that at which they are sold.  An interesting article on this new trend "Should You Overclock Your CPU?" is found at:
http://www.zdnet.com/computershopper/edit/smartshopper/buy_perspective/032299bp.html.
This web page has a good set of links to further information and tutorials on overclocking.  A great resource for overclockers is Overclockers.com.
2. Types of architecture (Review LM IIIB, section 5.A.):
1. CISC (_________(9) Instructional Set Computing) technology is the original architecture of all early microprocessors
2. RISC (_________(10) instruction Set Computing) technology, a newer approach than CISC.  It achieves higher performance via a simplified language manipulating a sophisticated architecture containing more registers than corresponding CISC chips. 
1. The Crusoe chip announced by Transmeta corporation on Jan., 2000 is a radically new approach to RISC microprocessor design. The key to this design is that, unlike traditional CPUs where all processing is done in hardware, the Crusoe's processing involves both hardware and software, thus reducing the size of the chip itself.  The following is an edited version of the descriptions on the Transmeta Web site.
1. The Crusoe architecture is composed of a compact hardware engine surrounded by a software layer. This eliminates the need for millions of transistors, replacing them with modifiable software.The first Crusoe chip consists of approximately one quarter of the transistors required for an all-hardware design withequivalent performanc e. This reduces power consumption and consequently heat thus Carusoe based computers will be easier to cool and, therefore, smaller and lighter.
2. Since the architecture is not associated with any O.S. such as Windows, the chip can exploit the best hardware innovations, without forcing users to change their software platform.
3. The chip's software can evolve separately from hardware. Users could download updated software from the Internet to improve chip performance without buying a new PC.  ...
With the Crusoe architecture, we now have powerful, energy-efficient family of chips that can operate all day and be easily fine-tuned to run a variety of software, on any operating system. This is remarkable. Moreover, using technologies such as the wireless Bluetooth innovation, these devices can be online constantly.
3. MMX: This is not a fundamental classification of architecture like CISC and RISC, but an enhancement of Intel/clone chips that have extra multimedia instructions built into the processor's instruction set.
4. EPIC (explicitly parallel instruction computing) is an emerging architecture that emphasizes parallel processing by using long, uniform instructions and numerous pipelines.  According to Boot magazine, "Intel claims that EPIC will surpass CISC and RISC and will take advantage of the hundreds of millions of transistors that will populate 21st century processors." EPIC is incorporated in the IA-64 (Intel Architecture-64), a 64 bit architecture implemented in the Itanium chip, introduced in 2001.
5. Multi-core processors are a special case of multiprocessor design that implement parallel processing architectures consisting of two or more full CPU cores within a single processor. (Review multiprocessing in LM IIIB, and see Multi-core computing from  Wikipedia.)
   
1. The first examples are called dual-core processors (DCP).  The following diagram compares the chip architectures of current dual-core architectures chips:

COMPARISON OF TYPES OF DUAL-CORE CHIP ARCHITECTURES

From "The Future of Microprocessors", ACM Queue, September, 2005. See the comparison of Intel's Xeon and AMD's Opteron.

2. Intel Plans to have an 80-core processor by 2010.  It will run at 3.1 GHz, delivering more than one teraflop of combined performance. The chip should make it to commercial production by 2010. "This kind of performance for the first time gives us the capability to imagine things like real-time video search or real-time speech translation from one language to another," Otellini said. Online-content providers such as Google and YouTube will likely require this level of processing power. By 2010, terra-scale servers will account for one quarter of all server sales by 2010, Intel projects. "We are talking about a fundamental change in the way that the whole computing infrastructure is built," said Intel CTO Justin Rattner. "At the core of that infrastructure will be the future data center, what we refer to as the mega data center."  Read this article, "Intel pushes for 80-core CPU by 2010" and/or view the video Intel envisions 80 cores coming to a server near you (video).
   

3. History and evolution of microprocessor technology:
1. See the interesting discussion of Moore's Law and the evolution of Intel microprocessors.
   
2. The history of the evolution of microprocessors is illustrated in Tech Encyclopedia at:
http://www.techweb.com/encyclopedia/defineterm?SHOWPIC=EVOLCHIP.GIF
4. The two dominant microprocessor families are the x86 processors manufactured by Intel, AMD, and Cyrix and the PowerPC processors manufactured by Motorola.  The different family members differ in architecture (word size, data bus size, address bus size, etc.). However, each different architecture, e.g. the Motorola G4, AMD Athlon, or the Intel Pentium IV, can have several "versions" differing in clock speed; generally, if the clock speed of a processor is doubled its speed is doubled .
1. Classification of  Intel microprocessors is shown in the following  Table from the Computer Desktop Encyclopedia. (A more detailed list is given in Figure 4-7 in the text.)



2. The newest generation of Intel and AMD chips,  with GHz clock speeds, make the preceding illustrations obsolete.
1. Intel is the dominant processor manufacturer; information about the state of the art Intel processors can be found on Intel's home page.
1. Information on Intel's Pentium 4 processor.
   
2. Intel is hyping a new, proprietary technology called "hyperthreading" (HT) which "enables a single processor to function as two "virtual" processors by executing two threads in parallel, allowing you and your software to multi-task more effectively than ever before."  For information on this see Intel's Hyper-Threading Technology.
3. Intel's emerging dual-core processor (DCP) technology, when combined with HT Technology, allows the Pentium D Processor to process four software threads simultaneously.  See the enlightening DCP vs HT and Intel Pentium D Processor Demo videos (click "Demo" links.)
2. AMD is Intel's main competitor; fortunately the "processor wars" between AMD and Intel fuel advances in CPU design and keep prices low. The AMD Athlon is the flagship of AMD processors; see the extensiver coverage in Wikipedia's Athlon article.  See the new 64 bit word processor architecture at AMD64 Platform.
   
3. PC Magazine emphasizes that benchmark results for computers incorporating the newest microprocessors are always very tentative and depend not only on the processor but other system components as well.  Also, one should remember that overclocked dual processor systems still have higher effective processing speeds.
3. For information on the PowerPC line of chips, previously found in Macs and still used in multiprocessor supercomptuers, {EXPAND?}
1. The PowerMac architecture is schematically presented (and compared to PC motherboards) in FIGURE PC-5C.
   
5. Additional Information  (links with strikethroughs are now unavailable (UGGGGG!); I am looking for alternatives.  If you find any, let me know on the class forum!):
1. For information on future Intel chips see: http://www.intel.com/home/scenes/stories/index.htm  
2. For a information about the latest chip technology and news see: http://www.cpu-central.com/
3. For PC Guides comparisons of Intel/Clone microprocessors (ONLY FOR HARD CORE TECHIES!) see http://www.pcguide.com/ref/cpu/fam/expl-c.html
4. For more information on (older) microprocessors click here.  (This helpful site has not been updated lately.  I don't know why.) 
   
5. For a detailed history microprocessors check out http://infopad.EECS.Berkeley.EDU/CIC/archive/cpu_history.html  

3. MICROCOMPUTER BUSES:

1. In modern PCs it is a combination of the local bus and multiple internal buses that connect the microprocessor to motherboard components (including expansion slots). Unfortunately there is a confusing inconsistency in the names of busses, especially "local bus", so double check the meaning of that word when reading other sources.  A schematic of a more modern representation of a micro system come from the article "Much A Do about CPUs" in Boot magazine, Jan., 1999.  (The name of Boot magazine has now been changed to Maximum PC.)  {UNFORTUNATELY, this informative diagram is no longer available.  I'm trying to find a replacement.  If anyone knows of or comes across an illustration of a PC buses, from a layperson's viewpoint, give me a reference on the forum. In the meantime, the PCI discussion in How Stuff Works will suffice.}
   
1. The microprocessor bus connects the different parts of of the processor, but, because it is part of the chip, is inaccessible to external devices.  It has to be interfaced to them via internal buses.
1. DMA (direct memory access) is a technology for transferring data from RAM directly to a device, e.g. a disk drive, without involving the CPU.  This allows much more efficient transfers of data between RAM and secondary storage than systems that must utilize the CPU to control this.
2. Ultra DMA is an improvement on DMA. The Ultra DMA/66 protocol transfers data in burst mode at a rate of 66.6 MBps, twice as fast as the previous DMA interface, Ultra DMA/33.
2. Internal (motherboard) buses which connect the microprocessor with other components of the motherboard, including expansion slots.
1. The system bus (also called the local bus) connects the microprocessor to the core-logic chipset and coordinates the data flow for each peripheral and device connected to the motherboard.
2. The backside bus connects the microprocessor directly to L2 cache.  It typically runs at the speed of the microprocessor bus itself (to optimize data transfers from cache to processor), and is, therefore, faster than the frontside (or local) bus.
3. The frontside bus (also called the system I/O bus) connects the core-logic chipset directly to motherboard components like primary memory, high speed peripherals and chips that link to internal buses.  When the higher speed VL-bus and PCI bus were introduced, they were called local buses, because they ran at the then-current speed of the local bus. Since then, local buses have gone beyond the speeds of VL-bus and PCI so they should be distinguished from the local bus (system bus) so they, along with other motherboard buses, are now considered to be part of the frontside bus.  (See section B, below and Figure PC-4.)
3. Expansion slots, connected to the frontside bus, allow peripheral devices to be connected to the microprocessor . It is necessary to insert an expansion card (hardware _________(11)) that is compatible with the particular frontside bus before one can connect the peripheral device. Third-party manufactures produce interfaces that allow their device to be connected to different computers makes.
3. External busses, e.g. SCSI ( parallel) and USB (serial) connect peripheral devices to the motherboard via hard-wired ports (without using expansion cards). One simply connects the compatible bus into the port and connects it to its device.  (See section C, below.)
4. PCMCIA (Personal Computer Memory Card International Association or "people can't memorize computer industry acronyms!") is the name of a new standard for expansion cards for portable computers. The PCMCIA card, officially called the PC Card, is credit card size and comes in three thicknesses that can house devises of varying sizes. See Figure PC-6B.
2. Standard motherboard buses are part of the frontside bus of a motherboard. (See typical current layout in Figure PC-5A future "hub architecture" in Figure PC-5B (from "How Data Moves Around Your Motherboard", MAXIMUM PC, Oct. 99) and a comparison in Figure PC-6.  The text's representationof internal buses is given in  Figure 3-38.):
1. Older IBM/clone buses, that have been superseded by the PCI bus, include
1. ISA, the original 8-bit Standard Architecture,  is now obsolete, but still appears on motherboards so that older devices may be connected and
2. three 32-bit buses: Micro Channel Architecture (MCA) utilized by IBM, Extended Industry Standard Architecture (EISA) used by many "IBM clone" makers, and VL-bus which rivaled the PCI bus but now survives only in older 486s.
2. PCI (Peripheral Component Interconnect) bus and its decendent PCI Express is a shared bus topology that facilitates communication among the different devices on the bus and provides "Plug and Play compatibility.  
1. "Plug and Play" means PCI cards (interfaces) of peripheral devices automatically configure themselves at startup. At the heart of PCI's design, built-in configuration registers and software will automatically keep track of every interrupt, coming from a peripheral device, being used in the system. When a new PCI peripheral is added, the PCI chipset will simply select an unused interrupt. No more changing jumpers. No more keeping track of IRQs.  Plug and Play is a feature of the new USB and FireWire peripheral buses; see C.c, below.
2. Most PCI systems will support three to five performance-critical peripherals. These peripherals will be either integrated directly onto the motherboard or can be added via PCI expansion cards, such as multimedia, graphics, disk drives and LAN cards.
3. PCI is designed to be synchronized with the clock speed of the microprocessor, e.g. the early 32 bit, 33MHz bus with a bandwidth of _________(12) Mbps. Later flavors of PCI included a 64-bit, 33MHz bus combination with a peak bandwidth of 264MB/s; a more recent 64-bit, 66MHz combination with a bandwidth of _________(13).
4. PCI uses all active paths to transmit both address and data signals, sending the address on one clock cycle and data on the next.
5. PCI has two formats: the 32 bit, 124-pin connection (the extra pins are for power supply and grounding) and a 64 bit, 188-pin expanded implementation.
6. PCI is designed to supplement, not replace, the traditional I/O busses so old ISA, EISA or MCA expansion cards can still be used when one switches to a new PCI-based PC.  In fact, PCI interfaces with ISA, EISA and MCA buses so expansion slots for these buses can co-exist with PCI slots on a motherboard.  See Figure PC-6.
7. PCI Express (a.k.a. PCIe) (in Wikipedia) is the latest incarnation of the PCI technology...See the Overview of PCI Express at Ars Technica.
1. PCI Express features point-to-point serial interconnections between devices and allows higher clockspeeds by carrying data in packets.
   
2. PCI Express is software compatible with existing PCI specifications.
For more detailed information on the PCI bus click here.
3. AGP (Accelerated Graphics Port) is a bus specification specially designed to speed up 3-D graphics on Intel based PCs.  The interface uses primary memory for refreshing the monitor screen and to support special effects required for 3-D graphics; however, when not being used for accelerated graphics, RAM is made available to the O.S. or applications.
1. AGP provides a direct connection between the display adapter and memory at 66 MHz, higher than the speed of the comparable PCI bus. A clock doubling technique boosts speed to 133 MHz.
2. AGP-equipped motherboards have only one AGP slot, which requires an AGP-compliant display adapter.
3. Intel has built AGP into a chipset for its Pentium II and subsequent microprocessors; the Pentium II was specially designed to work with the AGP chipset.
3. Standard external busses connect the motherboard system bus to other devices.
   
1. ATA  is the least expensive way to connect storage devices.  With the market introduction of Serial ATA (SATA) in 2003, the original ATA was retroactively renamed Parallel ATA (PATA).
1. Parallel ATA (also called "EIDE" or "IDE", which is an older standard) "ribbon" cables have limited lengths (about 36 inches, but usually less) because it transports data over parallel wires.   Therefore, they are normally used within the system unit of a PC, where It provides the most cost effective interface to secondary storage devices, especially hard disks.
2. Serial ATA is superceding PATA because it is more efficient (reduced voltage requirement) and less usceptible to interference (can thus have longer, thinner, more flexible cables with lower pin count).  They are still primarlily used to connect secondary storage devices to the motherboard, but it provids up to 300 MB/s data transfer rate.
2. SCSI (Small Computer Systems Interface) is the current preferred way to connect the widest range and number of devices via a single parallel port on a computer. (SCSI can be internal as well.) The original SCSI-1 is an 8-bit bus that allows up to seven devices to be attached per channel (port). Upgrades include SCSI-2 (8, 16 or 32 bit data paths) and SCSI (under development) whose goals are more devices on a bus (up to 32); faster data transfer; greater distances between devices (longer cables); more kinds of devices; and a structured protocol model.
3. Several new standards are evolving, but two plug-and-play daisy chaining (ability to chain devices together in a number of different ways without terminators or complicated set-up requirements) are rapidly replacing older ones.
1. USB (Universal Serial Bus), developed by IBM, Intel, and collaborators, is a "plug-and-play" peripheral bus for  daisy chaining I/O devices (and, perhaps, low-cost hard disk connections). USB allow seamless connections to 127 daisy chained devices per port.  Suitable high-speed devices include MPEG-2 video devices, plug-in telephones, data gloves, etc.  USB support is integrated into Windows 98 and above. The Web site of the nonprofit USB trade organization is www.usb.org.  (From Wikipedia) USB supports three data rates.
1. A Low Speed rate of 1.5 Mbit/s (183 KiB/s) that is mostly used for Human Interface Devices (HID) such as keyboards, mice, and joysticks.
2. A Full Speed rate of 12 Mbit/s (1.5 MiB/s). Full Speed was the fastest rate before the USB 2.0 specification and many devices fall back to Full Speed. Full Speed devices divide the USB bandwidth between them in a first-come first-served basis and it is not uncommon to run out of bandwidth with several isochronous devices. All USB Hubs support Full Speed.
3. A Hi-Speed rate of 480 Mbit/s (57 MiB/s).  Though Hi-Speed devices are commonly referred to as "USB 2.0", not all USB 2.0 devices are Hi-Speed. A USB device should specify the speed it will use by correct labeling on the box it came in or sometimes on the device itself.  (See Wikipedia's USB article for details.)
   
2. FireWire/IEE1394 is an evolving high-speed SCSI serial technology that is designed to support much higher data rates USB  (at least 400Mbps). (FireWire is Apple Computer's version of a new standard, IEEE 1394 High Performance Serial Bus.)  It has enough bandwidth to handle video I/O, live multimedia telecasts, DVD players, etc.; live multimedia applications currently require 100 - 200 Mbps and will need much more in the near future.   With its inexpensive design, high bandwidth, and ability to daisy chain up to 63 devices, it will soon be used for all types of peripheral devices.  In time, IEEE 1394 implementations are expected to replace and consolidate today's serial and parallel interfaces, including Centronic parallel, RS232-C, and SCSI.  Because IEEE 1394 is a peer-to-peer interface, one camcorder can dub to another without being plugged into a computer. With a computer equipped with the socket and bus capability, any device (for example, a video camera) can be plugged in while the computer is running.
4. External buses may be identified by the ports by which external devices are connected to those buses. See Figure 3-32 in the text.
4. Most computers have internal busses which have more bandwidth than their external busses (usually twice as much) since the bandwidth of the internal bus affects the speed of all operations and has less effect on the overall system cost than the bandwidth of the external bus.
5. For a comparison of bus bandwidths see the table from the article, "How Data Moves Around Your Motherboard", Maximum PC, Oct. 99

4.1 Yesterday's micro hardware uses today's software:

1. 8-bit and 16-bit hardware are obsolete!
2. 32-bit hardware has a large software base incorporating very sophisticated applications; however, these are rapidly being replaced by...
   
3. 64-bit and 128-bit hardware is being introduced.  High-end workstations, typically designed for graphics processing have 64-bit and 128-bit processors.

        With every new technological development, there follows a time-lag before applications appear that take advantage of it.  This is true of both hardware and software.  To account for this with new microprocessors, the instruction sets are made backwards compatible so that older software can run on computers based on the new architecture until new applications are produced. Also to minimize the time gap when new chips or operating systems are introduced, prototype processors and O.S. are sent to application developers so that they can make software available when the new technology is introduced.  Examples of such transitions include the following.

1. The IBM PS/2 line of computers which made the PC and MS-DOS obsolete, had a visual interface (the Application Manager) like the Mac. However comprehensive software that fully utilized the new operating system, OS/2, was "too little, to late" to overcome the popularity of Microsoft Windows, so the PS/2 is no longer available.
2. The introduction of Microsoft Windows 3.0 (a graphical user interface for MS-DOS), in the Spring of 1990 provided competition for OS/2. Windows 95, a completely new O.S. (not just a _________(14)) for micros was released by Microsoft in 1995 and its successors, _________(15) has been superseded by _________(16).
3. The Mac line of computers was introduces in 1984 and systematically upgraded since. Their operating system, current version System X, is based on POSIX, an advanced version of UNIX.  Mac software is extremely sophisticated, particularly in its multimedia features. However, even the old Macs have the friendly visual interface, so all of the applications bases on it have the productivity-increasing virtues that Microsoft emulate with their WINDOWS operating systems.
1. Apple at first committed to the PowerPC based Macs; during the transition from the 680x0 architecture the PowerPC micros ran 680x0 software via emulation, but Apple has now joined the Intel processor based systems. (See Section 5.C, below.)
2. Apple has introduced, in 1998, an innovative computer line with its new iMac, where the "i" stands for Internet.  It is a low-cost version of the Macintosh designed for extreme ease of use and transparent Internet access.
   
3. Apple abandoned the PowerPC platform in favor of Intel processors in its Mac Pro line of PCs.
   

A really neat Web site for comparison shopping for hardware is PRICE WATCH, whose URL is www.pricewatch.com/Check it out!  Another site that gives independent analysis and recommendations of hardware is C|NET's computers.com and shopper.com. For a peek at a "perfect 10" PC of today see the review of the Dell Dimension xps b800r in Maximum PC March 2000, p. 76.

5. TOTAL PORTABILITY, THE PERFECT WORLD OF THE FUTURE?:

1. In the past, computers of both IBM and Apple were more expensive than they could have been; the consumer has to pay for their research and development as well as advertising.
1. The IBM clones often had better price/performance ratios than the IBM systems and, in many cases, simply had better performance (e.g. Compac was considered by many to have been the pioneering company producing Intel based microcomputers, but it has been replaced by companies like Dell and Gateway).
2. Except for a few years during the mid-nineties, there have been no Mac clones because Apple refused to license its operating system. This  changed for a while and companies like Power Computing produced by Mac clones, but Apple has reverted to its former no-clone position.
2. A computer can be made to "emulate" another computer, i.e. run the O.S. and applications of another computer. This makes it possible to have a computer that can run any program.   Emulation has two forms:
1. Software emulation is accomplished completely in software. Because the microprocessor must translate the instruction set of the "foreign" computer into its "native" instruction set, the emulation is slower than real computer performance.
2. Hardware emulation is accomplished by actually using the same CPU as that on the foreign micro. If the two computers have the same microprocessor, the emulation is straight-forward, i.e. a "plug-in" board containing the "foreign" computer's ROM chips and emulation software may be all that is needed.
3. EXAMPLES:
1. Macs have optional plug-in boards as well as software emulators which allows them to run any Windows software. In mid-1998 a software Mac emulator for Windows was introduced.
2. Two popular computers of the late 80's and early 90's. the Atari ST and Commodore Amiga, both had hardware and software emulators that allow them to run MS-DOS software. However, these fine computers have been marketed out of existence!  The ST and Amiga computers were capable of modification that allow them to run both DOS/Windows and Mac software. In fact, the Commodore Amiga had a "universal emulator" called "Emplant" that was designed to emulate any other micro whose ROM chips could be inserted into the emulator board, or whose contents could be down loaded into RAM.  Unfortunately both Atari and Amiga, despite their innovative features, have been run out of business by the marketing power of Microsoft, PC clones, and Apple.
   
3. "What's in a name? A rose by any other name..." Unfortunately this is not true in the PC market; the marketing giants like Microsoft and Intel consistently squeeze better technologies out of market share and maybe out of business. Many believe the "Wintel" (Windows/Intel) machines (Windows software running PCs with Intel processors) will form a monopoly that will stifle innovation.) For example, the Amiga and STs, once the most computing power for the buck, are now virtually nonexistent, although they survived for a while in Europe.
4. Lindows is an operating system strategy that promises to unite the Linux and Windows platforms.
4. THE BIG QUESTION: Ideally, we should be able to run any type of software on any computer i.e. applications should be completely portable. However, this would nullify Microsoft's virtual monopoly on O.S. so that software giant is not only not supporting the effort but is doing subtle things (like giving free Microsoft-specific alternatives) to fragment the effort. New programming languages like Java and technologies like the NC all focus on such an ideal situation.  Consequently, only time will tell, if this will ever actually come about!  Even if it does not, there are undoubted on-going benefits for the consumer in that the competition holds the price tag down and accelerates innovation. On the other hand, too often buggy prototypes are rushed to market; these can lead to very frustrating experiences!

1. The performance distinction between microcomputers, minicomputers, and mainframes is hazy and transient; the Pentium, AMD, and PowerPC based "micros" are all more powerful than the mainframes of a decade ago and the "minis" of a few years ago. The important distinction is between "personal computers" ("PCs"), whose microprocessor the user can have all to himself (if that is desired) and "multiuser computers" ("MCs") which are, by design, a single processor or multiprocessor complex shared by numerous dumb terminals. (It should be noted that the micros with 80386 or 68030 processors (or above) are capable of supporting multiple users, so it is how they are used in a computer's design that makes them either PCs or MCs!)
2. The alternative to a MC is a distributed computing system.   In a multiuser system, if the central computer "goes down" every user is out of luck; in a distributed computing environment when a computer malfunctions only the user of that computer is effected.  (See Figure PC-7 for a comparison of distributed computer systems versus the PC.)  Three versions of distributed PC systems are:
1. The "_________(17) Computers" (NCs as opposed to PCs) are being based on the idea that "the _________(18) IS the computer!
2. Networked workstations, e.g. Windows NT workstations, are PCs that are interconnected as well as connected to printers, servers (e.g. file servers which are computers whose hard disk is accessible to everyone in the network), net modems, etc.
3. NetPCs are stripped down PCs (but containing local secondary storage) designed specifically to be part of a network via which they access data, application software, etc.  Their locally stored software are installed, maintained, and updated, via the network, under centralized control.
The distributed computing concept is personified by Suns' moto, "The computer IS the NETWORK" and is the foundation of Microsoft's ".Net strategy".   Networking is discussed in detail, later in the course, in Learning Module VI, Networking and Computers.
3. The ideal, ultimate goal is an optimally distributed computing environment within which computing power/facilities is spread efficiently (and dynamically!) over the user base. This is rapidly approaching with the advances involving the Internet and high-speed communications hardware combined with super computer access.  Wireless access to the Internet means that we can have limitless resources available from anywhere!

1. MICROCOMPUTER COMPONENTS (See Fig. PC-1A):
1. In general a microcomputer consists of:
1. The System Unit case houses the motherboard, daughterboards, expansion cards, power supply, internal secondary storage devices, and CD-ROM devices.
2. Peripheral Devices, for input, output, secondary storage, and communications.
2. Architecture and Organization of a typical Micro System:
1. A schematic of a generic microcomputer is given in Figure PC-3shows components that communicate via computer buses.
2. In general, a computer bus is a set of conductors that provides a common pathway between multiple electronic chips.
1. Buses typically data, addresses, and control signals.
2. Bus bandwidth of a bus is the number of megabytes it can transmit per second.
3. Microprocessor and primary memory communicate with the motherboard buses via the core-logic chipset.
4. Peripheral devices are connected, via external busses, to computer ports, which are, in turn, connected to the motherboard via hardware interfaces that are either serial or parallel.
2. MICROPROCESSOR ARCHITECTURE:
1. MPs, differing greatly in architecture, can be loosely classified according to their speed which depends on word size, data bus size, address bus size, and clock speed.
2. Types of architecture (Review LM IIIB, section 5.A.):
1. CISC, the original architecture of all early microprocessors
2. RISC achieves higher performance via a simplified language manipulating a sophisticated architecture containing more registers than corresponding CISC chips.
3. EPIC, an emerging architecture, emphasizes parallel processing using long, uniform instructions and numerous pipelines.

The Crusoe, from Transmeta corporation, is a radically new approach to RISC microprocessor design that utlizes software as well as hardware in its processing, thus allowing the chip size and its power consumption to be reduced.  This makes it ideal for portable computers.
4. MMX is a multimedia enhancement of the processor's instruction set.
3. The two dominant microprocessor families are the x86 processors manufactured by Intel, AMD, and Cyrix and the PowerPC processors manufactured by Motorola. All of these have several "versions" differing in clock speed.
1. Classification of  Intel microprocessors is shown in the  Table from the Computer Desktop Encyclopedia:
1. Intel and AMD dominate the market for Windows based microcomputers, but the relatively unknown 64-bit RISC Alpha processor design is considered by many to be superior.
2. The PowerPC line of chips is used in Macs.
3. MICROCOMPUTER BUSES:
1. The microprocessor is connected to the motherboard by acombination of the local bus and multiple internal buses.
1. The microprocessor bus, which interconnects the different parts of of the microprocessor, is interfaced to external devices via internal buses.  DMA and Ultra DMA allow data to be transferred from RAM directly to a device, without involving the CPU.
2. Internal (motherboard) buses connect the microprocessor with other motherboard components of the motherboard.  These include:
1. the system bus (also called the local bus) which connects the microprocessor,
2. the backside bus which connects the microprocessor directly to L2 cache, and
3. the frontside bus (also called the system I/O bus) which connects the core-logic chipset directly to motherboard components.
1. Expansion slots, connected to the frontside bus, allow peripheral devices to be connected, via expansion cards, to the microprocessor .
3. External busses, e.g. SCSI ( parallel) and USB (serial) connect peripheral devices to the motherboard via hard-wired ports (without using expansion cards).
4. PCMCIA  is the standard for expansion cards for portable computers.
2. Standard Internal buses are part of the frontside bus of a motherboard.
1. Older IBM/clone buses (now superseded by the PCI bus) include ISA, MCA, EISA, and VL-bus.
2. PCI bus, now the standard in virtually all micros, is a modern 32 bit, 33MHz bus with a bandwidth of 32x33 Mbps and "Plug and Play compatibility.  This bus is designed to supplement, not replace, the traditional I/O busses so old ISA, EISA or MCA expansion cards can still be used when one switches to a new PCI-based PC.
3. AGP (Accelerated Graphics Port) is a bus specification specially designed to speed up 3-D graphics on Intel based PCs by providing a direct connection between the display adapter and memory.
3. Standard external busses:
1. EIDE, the oldest (and least expensive) way to connect external storage devices, is being superseded by SCSI, USB and FireWire.
2. SCSI, currently the most popular way to connect devices via a single parallel port on a computer, has three versions, SCSI-1 (an 8-bit bus that allows up to seven devices to be attached per port), SCSI-2 (8, 16 or 32 bit data paths) and SCSI which is under development.
3. USB a "plug-and-play" peripheral bus for daisy chaining up to 127 I/O devices per port.
4. FireWire/IEE1394 supports support much higher data rates than USB  (at least 400Mbps). and can daisy chain up to 63 devices per port.
4. THE HARDWARE-SOFTWARE DILEMMA:
1. Yesterday's micro hardware uses today's software, but
2. Today's micro hardware awaits tomorrow's software, so
do you buy state-of-the-art hardware or the computer with the most software?  (The "best bang for the buck" are systems that are one generation older than the state of the art.)
5. TOTAL PORTABILITY, THE PERFECT WORLD OF THE FUTURE, means that all computers would be independent of specific processors and operating systems.
6. DISTRIBUTED COMPUTING, THE FUTURE OF COMPUTING, means that future computer users are not limited by their "personal" computer because it has access to all the resource (including other computers like supercomputers) of the network to which it is connected.  This is personified by Suns' moto, "The computer IS the NETWORK".  Wireless access to the Internet means that we can have limitless resources available from anywhere!

FIGURE PC-2 Current Motherboard (Asus - P4P800 Deluxe)

Side View of Back Panel Connector

Pictures from TigerDirect.com

 Compare  and Figures PC-2 with the older, generic motherboard in Figure 4-3 of Shelly & Cashman.

FIGURE PC-4:  BLOCK DIAGRAMS OF THE CPU & MOTHERBOARD

CPU Block Diagram

Motherboard Block Diagram

 

FIGURE PC-5A

FIGURE PC-5B

FIGURE PC-5C PowerMac Architecture 1-2. Dual Processors and frontside bus 3. AGP Graphics Card 4. System Controller 5. DDR SDRAM 6. PCI-X controller 7. I/O subsystems hub (ATA, Eternet, Firewire, USB) 8. Serial ATA hard drives

FIGURE PC-6

 
 

FIGURE PC-7: DISTRIBUTED COMPUTING SYSTEMS VS. THE PC