PC Hardware in a Nutshell | O'Reilly Media

Content preview·Buy reprint rights for this chapter

This chapter covers a mixed bag of important fundamental information about PCs, including how PCs are defined, an overview of PC components and technologies, a brief explanation of system resources, guidelines for building, buying, and upgrading PCs, smart buying practices, and suggestions as to what to do with old PCs.

Who decides what is and what is not a PC? That question is not as trivial as it sounds, because there has never been (and probably will never be) an all-embracing de jure standard to define the PC. IBM created the de facto PC standard (and trademarked the name) when they shipped the first IBM Personal Computer in 1981. For more than five years, until their introduction of the ill-fated proprietary PS/2 line in 1987, IBM defined the PC standard. For a short time thereafter, some considered that Compaq defined the standard. But the days when any PC maker defined the PC standard are far in the past.

These days, Intel and Microsoft jointly define the de facto PC standard. In fact, a good working definition of a PC is a computer that uses an Intel or compatible processor and can run a Microsoft operating system. Any computer that meets both requirements—a so-called Wintel computer—is a PC. A computer that does not is not. Computers based on some Intel processors cannot run any Microsoft operating system, and thus are not PCs. Conversely, some computers with non-Intel processors can run Microsoft operating systems, but also do not qualify as PCs. For example, DEC Alpha minicomputers running Windows NT 4 are not PCs.

Two formal documents, described in the following sections, define the joint Intel/Microsoft standards for systems and components you are likely to be working with. These standards are de facto in the sense that system and peripheral makers are not required to comply with them to manufacture and sell their products. They might as well be de jure standards, however, because compliance is required to achieve such nearly mandatory certifications as inclusion on the Windows NT/2000/XP Hardware Compatibility Lists.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Who decides what is and what is not a PC? That question is not as trivial as it sounds, because there has never been (and probably will never be) an all-embracing de jure standard to define the PC. IBM created the de facto PC standard (and trademarked the name) when they shipped the first IBM Personal Computer in 1981. For more than five years, until their introduction of the ill-fated proprietary PS/2 line in 1987, IBM defined the PC standard. For a short time thereafter, some considered that Compaq defined the standard. But the days when any PC maker defined the PC standard are far in the past.

These days, Intel and Microsoft jointly define the de facto PC standard. In fact, a good working definition of a PC is a computer that uses an Intel or compatible processor and can run a Microsoft operating system. Any computer that meets both requirements—a so-called Wintel computer—is a PC. A computer that does not is not. Computers based on some Intel processors cannot run any Microsoft operating system, and thus are not PCs. Conversely, some computers with non-Intel processors can run Microsoft operating systems, but also do not qualify as PCs. For example, DEC Alpha minicomputers running Windows NT 4 are not PCs.

Two formal documents, described in the following sections, define the joint Intel/Microsoft standards for systems and components you are likely to be working with. These standards are de facto in the sense that system and peripheral makers are not required to comply with them to manufacture and sell their products. They might as well be de jure standards, however, because compliance is required to achieve such nearly mandatory certifications as inclusion on the Windows NT/2000/XP Hardware Compatibility Lists.

PC 99 System Design Guide (PC 99) is a book-length document that defines required, recommended, and optional (neither required nor recommended, but must meet the standard if included) characteristics for several classes of PCs, including

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

The following sections provide a quick overview of the components and technologies used in modern PCs.

One of the great strengths of the PC architecture is that it is extensible, allowing a great variety of components to be added, and thereby permitting the PC to perform functions its designers may never have envisioned. However, most PCs include a more-or-less standard set of components, including the following.

Motherboard

The motherboard , described in Chapter 3, is the heart of a PC. It serves as "Command Central" to coordinate the activities of the system. Its type largely determines system capabilities. Motherboards include the following components:

Chipset: The chipset provides the intelligence of the motherboard, and determines which processors, memory, and other components the motherboard can use. Most chipsets are divided physically and logically into two components. The Northbridge controls cache and main memory and manages the host bus and PCI expansion bus (the various buses used in PCs are described in Chapter 3). The Southbridge manages the ISA bus, bridges the PCI and ISA buses, and incorporates a Super I/O controller, which provides serial and parallel ports, the IDE interface, and other I/O functions. Some recent chipsets, notably models from Intel, no longer use the old Northbridge/Southbridge terminology although the functionality and division of tasks are similar. Other recent chipsets put all functions on one physical chip.
CPU slot(s) and/or socket(s): The type of CPU slot or socket determines which processors the motherboard can use. The most popular CPU connectors are Socket 370 (current Intel Pentium III and Celeron processors), Socket A (current AMD Athlon and Duron), Socket 478 (current Pentium 4), Socket 423 (old-style Pentium 4), Slot 1 (old-style Pentium II/III and Celeron), Slot A (older-style Athlon), and the nearly obsolete Socket 7 (Intel Pentium and AMD K6-* processors). Some motherboards have two or more CPU connectors, allowing them to support multiple processors. A few motherboards have both Slot 1 and Socket 370 connectors, allowing them to support either type of CPU (but not both at once).

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

PCs have four types of system resources—Interrupt Request lines, DMA channels, I/O ports, and memory ranges. Many system components and peripherals require one or more of these resources, which raises the twin problems of resource availability and resource conflicts. Resource availability is particularly important with regard to IRQs, which are in high demand, and of which only 16 exist. Resource conflicts can occur when two devices are assigned the same resource, in which case one or both devices may not function or may function unpredictably. Resource conflicts may occur even with plentiful resources, such as I/O ports, where many are available and only a few are in use.

A frequent cause of problems when building or upgrading PCs is a shortage of required resources or unintentional resource conflicts that occur when a new component is installed that was inadvertently configured to use a resource already in use. Two technologies, PCI and Plug-N-Play, used in conjunction with recent versions of Microsoft operating systems (Windows 95 OSR2, Windows 98, and Windows 2000/XP) go a long way toward extending the availability of resources and preventing conflicts. Even in such an ideal environment, however, resource conflicts sometimes occur, particularly if you are using older "legacy" hardware. The following sections describe what you need to know about PC resources and how to manage them.

When a component or peripheral, such as a network adapter or sound card, needs to get the CPU's attention, it does so by generating a signal on an Interrupt ReQuest Line (IRQ). Table 1-2 lists IRQs and the devices that typically use them.

Table 1-2: 8/16/32-bit ISA/PCI standard IRQ assignments
IRQ	Bus type

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

The make-or-buy decision is a fundamental business school concept. Does it make more sense to make a particular item yourself or just to buy it? With entry-level PCs selling for $600 and fully equipped mainstream PCs for $1,500, you might wonder why anyone would bother to build a PC. After all, you can't save any money building one, can you? Yes, you can. Quite a bit of money, in fact. But that's not the only reason to build a PC. Here are good reasons to do so:

Choice: When you buy a PC, you get a cookie-cutter computer. You may be able to choose such options as a larger hard drive, more memory, or a better monitor, but basically you get what the vendor decides to give you. And what you get is a matter of chance. High-volume direct vendors like Gateway and Dell often use multiple sources for components. Two supposedly identical systems ordered the same day may contain significantly different components, including such important differences as motherboards or monitors with the same model number but made by different manufacturers. When you build a PC, you decide exactly what goes into it.
Component quality: Many computer vendors save money by using OEM versions of popular components. These may be identical to the retail version of that component, differing only in packaging. But OEM versions have several drawbacks. Many component vendors do not support OEM versions directly, instead referring you to the computer vendor. And OEM versions often differ significantly from the retail-boxed version. For example, Micron used the popular Intel SE440BX Seattle motherboard in many of its systems, but modified the Intel-supplied BIOS. That means owners of those systems cannot use Intel BIOS updates. Instead, they must depend on Micron to provide an updated BIOS. Dell and other major makers sometimes use downgraded versions of popular products, for example, a big-name video card that runs at a lower clock rate than the retail version. This allows them to pay less for components and still gain the cachet from using the name-brand product.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Sometimes it's sensible to upgrade a PC. Other times, it's not. Whether it is economically feasible to upgrade a particular PC depends largely on how old the PC is, its existing configuration, and what you expect it to do.

PCs less than a year or two old are usually easy to upgrade. Components are readily available and sell at market prices. Necessary BIOS upgrades and firmware revisions are easy to obtain. PCs more than two or three years old are harder and more expensive to upgrade. Necessary components, particularly memory and BIOS upgrades, may be difficult or impossible to obtain. Even if you can obtain them, they may be unreasonably expensive. Upgrading one item often uncovers a serious bottleneck elsewhere, and so on. In general, restrict older PCs—say anything older than a Pentium III or late-model Celeron system—to minor upgrades such as adding memory or replacing a hard disk. Although you can do significant upgrades on older systems, it seldom makes economic sense to do so. If an older PC requires more than minor upgrades to meet your expected needs for the next year or so, it's probably not a good upgrade candidate.

The reason most people upgrade their PCs is to improve performance. The good news is that there are several relatively inexpensive upgrades that may yield noticeable performance increases. The bad news is that some are easier than others, and that doing all of them can easily cost as much or more than simply buying a new PC.

Processor: Upgrading the processor improves overall system performance. In general, upgrade only within the same generation, for example, a Pentium II/266 to a Celeron/500, or a Pentium 133 to a K6-2/400. Avoid upgrade kits that jump generations, e.g., kits to upgrade a 486 to Pentium, or a Pentium to Pentium II. These kits are usually expensive, provide limited performance improvements, and are often plagued with compatibility problems. If you upgrade within the same generation, you may have to upgrade your BIOS at the same time (usually a free download), and you may have to buy an adapter (for example, to install a Socket 370 Celeron in an older Slot 1 motherboard). In general, newer systems are easy to upgrade, and older systems are more difficult (or impossible) to upgrade. Upgrading processors is covered in Chapter 4. Cost: $30 to $200 (although you can spend much more). Difficulty: Easy to difficult, depending on the system and the processor.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

A decade ago, most computer products were bought in retail computer stores. Retail sales still make up a significant chunk of computer product sales—although the emphasis has shifted from computer specialty stores to mass-market resellers like Best Buy and Costco—but the majority of computer products are now bought from direct resellers, via toll-free telephone number or the Web. Local brick-and-mortar retailers, with their high overheads, simply cannot match direct reseller prices and stay in business. Nor can they match direct reseller companies for breadth of selection or convenience. We frequently order components late in the evening. Early the next morning, our FedEx guy drops them on the front porch. All without our having to leave the house.

That said, there are some drawbacks to buying from direct resellers. You're dealing with an anonymous company, probably located far away. You must know exactly what you want, and you need to understand the pitfalls of dealing with direct resellers. Most direct resellers are reputable, but some are not. Even reputable resellers differ greatly in their business practices, so it's important to understand the rules before you start playing the game. We've bought hundreds of thousands of dollars worth of products from direct vendors over the last decade or so, and have learned some things from that experience. Here are some guidelines to keep in mind:

Research the product: Make sure you know exactly what you're buying before you order it. For example, a hard disk may be available in two versions, each with the same model number but with sub-model numbers to designate different amounts of cache. Or you may find that a given hard disk maker manufactures two models of the same size that differ in both price and performance. Always compare using the exact manufacturer model number. Before you buy a product, research it on the manufacturer's web site and on the numerous independent web sites devoted to reviews. We use

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

So what do you do with an old PC that would cost too much to upgrade to current standards? We encounter that question frequently around here. We have everything from the latest multiprocessor boxes to creaking 386s. In fact, our original 1984-vintage IBM PC/XT died only a couple years ago, and it was doing useful work until its untimely demise. Here, in no particular order, are ten useful things to do with an old PC:

Give it to your spouse: In many households, one spouse is a PC power user and the other is much less demanding. She works at home doing serious number crunching and plays Quake for relaxation, while he just checks his email periodically and uses the Web to keep up with the PGA Tour results. Or viceversa. He might be happier having an older system all to himself than he would be sharing the latest, fastest PC. While you're at it, consider installing a home network, if only to share your Internet connection. You can do so using a traditional wired Ethernet, 802.11 wireless networking, or even Home Phone Line Alliance (HPNA) networking. The cost can be as little as $50 for a couple of decent Ethernet cards and a 100BaseT cable.
Give it to your kids: Younger kids want to play educational games, some of which require a lot of PC, but many of which run just fine on a two- or three-year-old system. Older kids need word processing, web browsing, and email, but may also want to run games, some of which are quite demanding. Before you pass the old system on to the kids, consider doing one or more "$50 upgrades"—$50 for a faster processor, $50 to add RAM, $50 for a new video card, and, if necessary, $50 to replace the CD-ROM drive with a DVD-ROM drive. Before you do much more than that, remember that you can buy or build a pretty competent PC nowadays for $500 or thereabouts, not including the monitor.
Give it to an elderly neighbor or relative

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Popping the lid of a PC for the first time can be pretty intimidating, but there's really no need for concern. There's nothing inside that will hurt you, other than sharp edges and those devilish solder points. There's also nothing inside that you're likely to damage, assuming you take the few simple precautions detailed in this chapter.

Some PCs—particularly those from office supply and electronics superstores—have seals that warn you the warranty is void if they're broken. This isn't so much to protect them against your ham-handedness as it is to ensure that you have to come back to them and pay their price for upgrades. We advise friends and clients to break such seals if they need to, do their own upgrades, and fight it out later if they have a problem that should be covered under warranty.

We've never heard of anyone being refused warranty service because of a broken seal, but there's always a first time. If you have a sealed PC that is still under warranty, the decision is yours. Note that hard disks are a special case. Breaking the seal on a hard disk does actually destroy it and will without question void the warranty.

Those issues aside, feel free to open your PC and tinker with it as you see fit. Far from forbidding you from working on your own PC, most mail-order and retail computer vendors actually expect you to do your own upgrades. As a matter of fact, most of them will try to talk you into doing your own warranty repairs so that they can avoid sending a technician to do them for you. The rest of this chapter explains the fundamentals you need to understand to start upgrading and repairing your PC.

We've repaired, upgraded, and built hundreds of systems over the years, and learned a lot of lessons the hard way while doing it. Here are the rules we live by, some big, some small, and some more honored in the breach. We'll admit that we don't always take each of these steps when we're doing something simple like swapping a video card, but you won't go far wrong following them slavishly until you have enough experience to know when it's safe to depart from them.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

We've repaired, upgraded, and built hundreds of systems over the years, and learned a lot of lessons the hard way while doing it. Here are the rules we live by, some big, some small, and some more honored in the breach. We'll admit that we don't always take each of these steps when we're doing something simple like swapping a video card, but you won't go far wrong following them slavishly until you have enough experience to know when it's safe to depart from them.

Back everything up: Twice. Do a verify pass, if necessary, to make sure that what is on the backup tape matches what is on the disk drive. If you're connected to a network, copy at least your data and configuration files to a network drive. It's much easier to retrieve them from there than it is to recover from tape. If there's room on the network drive, create a temporary folder and copy the entire contents of the hard disk of the machine about to undergo surgery. If you have neither a tape drive nor a network volume, but you do have a CD or DVD writer, back up at least your important data and configuration files to optical discs. About 99 times in a hundred all of this will be wasted effort. The hundredth time—when everything that can go wrong does go wrong—will pay you back in spades for the other 99.

If you don't have a tape drive or a CD/DVD writer, installing one is an excellent first upgrade project. Floppy disks just aren't good enough for backup nowadays.
Make sure you have everything you need before you start: Have all of the hardware, software, and tools you'll need lined up and waiting. You don't want to have to stop in mid-upgrade to go off in search of a small Phillips screwdriver or to drive to the store to buy a cable. Our first rule of upgrading says you won't find the screwdriver you need and the store will be closed. Make sure you have a boot disk with drivers for your CD-ROM drive, and test it before you start tearing things down. Create a new emergency repair diskette immediately before you start the upgrade. Make certain you have the distribution disks for the operating system, backup software, and any special drivers you need. If you're tearing down your only PC, download any drivers you will need, copy them to floppies, and unzip them if necessary

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

It's worthwhile to assemble a toolkit that contains the hand tools and software utilities that you need to work on PCs. If you work on PCs only occasionally, you can get by with a fairly Spartan set of tools. If you work on PCs frequently, devote some time, effort, and money to assembling a reasonably complete set of hand tools and utilities. The following sections detail the components that we've found worth carrying in our toolkits.

You don't need many tools for routine PC upgrades and repairs. We've successfully repaired PCs using only a Swiss army knife, but a more complete set of tools makes jobs easier. Putting together a dedicated PC toolkit and keeping it in a fixed location avoids the hassle of looking for a tool when what you really want to do is work on your PC.

Your first thought may be to buy one of those PC toolkits available from various sources, but we suggest you avoid them. Inexpensive kits available from most mail-order vendors contain shoddy tools and are not worth even their low price. Kits available from specialty catalogs like Specialized Products (http://www.specializedproducts.com) and Jensen (http://www.jensentools.com) are fine if you fix PCs for a living (and if your company buys the kit). Otherwise they're overkill and much too expensive.

Instead of buying any of the prepackaged kits, head for Sears and assemble your own PC toolkit. The basic tools you need for routine PC work cost under $50. Store these tools together, using a tool wrapper (available from auto parts stores) or a zipper case (available from specialty tool vendors or a home improvement warehouse). You can often buy sets of pliers, screwdrivers, and so forth for less than buying them each individually. We carry only the tools we need, so we usually buy the set, remove the ones we really wanted for our toolkit, and contribute the remainder to the general stock of tools around the house. Table 2-1 lists what we carry and recommend as a basic kit, with Sears part numbers in parentheses.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

After you assemble a toolkit with the hand tools and utilities described in the preceding sections, you have everything you need to upgrade or repair a PC except the new components. Before you get started, take a few minutes to read through the following sections, which describe the common procedures and general knowledge you need to work on PCs. These sections describe the general tasks you perform almost any time you work on a computer—things like opening the case, setting jumpers and switches, manipulating cables, and adding or removing expansion cards. Instructions for specific tasks like replacing a motherboard, disk drive, or power supply are given in the relevant chapter.

Although you may be raring to get in there and fix something, taking the time to prepare properly before you jump in pays big dividends later. Before you open the case, do the following:

Make sure it's not a software problem: The old saying, "If all you have is a hammer, everything looks like a nail," is nowhere more true than with PC repairs. Just as surgeons are often accused of being too ready to cut, PC technicians are always too ready to pop the lid. Before you assume that hardware is causing the problem, make sure the problem isn't being caused by an application, by Windows itself, or by a virus. Use your hardware diagnostic utility and virus scanner before you assume the hardware is at fault and start disconnecting things.
Think things through: Inexperienced technicians dive in willy-nilly without thinking things through first. Experienced ones first decide what is the most likely cause of the problem, what can be done to resolve it, in what order they should approach the repair, and what they'll need to complete it. Medical students have a saying, "When you hear thundering hooves, don't think about zebras." In context, that means that you should decide the most likely causes of the problem in approximate ranked order, decide which are easy to check for, and then eliminate the easy ones first. In order, check easy/likely, easy/unlikely, hard/likely, and finally hard/unlikely. Otherwise, you may find yourself tearing down a PC and removing the video card before you notice that someone unplugged the monitor.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

The motherboard is the heart of a PC. Some manufacturers use the terms system board , planar board , baseboard , or main board , and Intel calls their motherboards desktop boards . No matter what you call it, the motherboard defines the PC. It provides the common link to all other components inside the PC, including the CPU, memory, disk drives, video and sound adapters, keyboard, mouse, and other peripheral components. If you are building a PC, choosing the motherboard is the most important decision you make and can be one of the most difficult. If you are upgrading a PC, replacing the motherboard is often the best and most cost-efficient means of doing so. If you are buying a PC, the motherboard it uses determines its functionality and future upgradability. This chapter describes the characteristics of motherboards, provides purchasing guidelines, and explains how to install and configure a motherboard.

Several characteristics differentiate motherboards, including physical characteristics, which in combination are called the form factor; the chipset used, which defines the capabilities of the motherboard; the processors the motherboard supports; the BIOS it uses; and the internal and expansion buses that it supports. The following sections examine each of these factors.

Motherboards differ in size, shape, position of mounting holes, power supply connector type, and port types and locations. Together, these differences define the form factor of the motherboard. Form factor is a critical issue when you upgrade a system, because the replacement motherboard must physically fit the case and use the existing power supply connectors. Form factor doesn't matter when you're building a new PC. You simply select the best motherboard for your needs, and then buy a case that fits it. Many motherboard manufacturers build similar motherboards in different form factors. Here are the form factors you may encounter:

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Several characteristics differentiate motherboards, including physical characteristics, which in combination are called the form factor; the chipset used, which defines the capabilities of the motherboard; the processors the motherboard supports; the BIOS it uses; and the internal and expansion buses that it supports. The following sections examine each of these factors.

Motherboards differ in size, shape, position of mounting holes, power supply connector type, and port types and locations. Together, these differences define the form factor of the motherboard. Form factor is a critical issue when you upgrade a system, because the replacement motherboard must physically fit the case and use the existing power supply connectors. Form factor doesn't matter when you're building a new PC. You simply select the best motherboard for your needs, and then buy a case that fits it. Many motherboard manufacturers build similar motherboards in different form factors. Here are the form factors you may encounter:

AT, Baby AT (BAT), and LPX: All of these form factors are based on the motherboard used in the original 1984 IBM PC AT and are now obsolescent or, in the case of LPX, obsolete. Although it is still possible to buy AT and BAT motherboards, selection is limited (e.g., Athlon and Pentium 4 motherboards are not available in these form factors) and most AT and BAT motherboards do not include the latest chipsets and other technologies. The only reason to buy an AT or BAT motherboard is to upgrade an existing system without replacing the power supply and case. Good cases with power supply are available for $65 or so, and the power supply in an older case is likely to be near the end of its life, so we do not recommend buying a motherboard in any of these form factors.
ATX and variants

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

You can sometimes upgrade a system cost-effectively without replacing the motherboard. The more recent the system, the more likely this is true. The easiest upgrade is always replacing a processor with a faster version of the same processor. Doing that may simply mean pulling the old processor and replacing it with the faster one, although a BIOS upgrade may also be needed. Alas, there is no guarantee that a given motherboard will support a faster version of the same processor or that a required BIOS upgrade will be available, and the rapid advances in processors means that a faster version of the same processor may no longer be available.

The next-easiest upgrade is to replace the processor with a later model from the same generation. For example, you may be able to replace a Pentium II/350 with a Pentium III/850 or, by using a slocket adapter, with a cheap, fast Celeron. When upgrading to a later-model processor, a BIOS upgrade will almost always be needed, and you should check the motherboard manufacturer's web site carefully to determine which configurations are supported.

Faster processors may draw more current, and the VRMs on an older motherboard may be inadequate to support the new processor. Even if the processor appears to work properly at first, running it for a while may damage both the motherboard and the processor. Always check to make sure that the exact processor you plan to install is supported by the motherboard.

It doesn't make sense to stretch an old motherboard too far. Just because you can upgrade a system without replacing the motherboard doesn't mean that you should do so. Motherboards are inexpensive, typically $75 to $150. Doing an in-place upgrade instead of replacing the motherboard leaves you with the limitations of the old motherboard and may limit the performance of the new processor. Before you decide to keep the motherboard, find out the costs and benefits of replacing it instead. Don't forget to factor in the supplementary benefits of a new motherboard—a better chipset and BIOS, support for the latest hard disk standards, etc. You may well decide it's worth spending the money to replace the motherboard. In fact, you may decide simply to retire the existing system to less demanding uses and build a new system.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Installing a motherboard for the first time intimidates most people, but it's really pretty easy if you do it by the numbers. Before you get started, prepare a well-lighted working area, ideally one with all-around access. The kitchen table (appropriately protected) usually works well. Have all tools and parts organized and ready to go. Open the box of each new component, verify contents against the manual or packing list to make sure no parts are missing, examine the components to ensure they appear undamaged, and do at least a quick read-through of the manual to familiarize yourself with the products.

See also Chapter 28, for photographs of this process.

If you are replacing a motherboard, you must remove the old motherboard before installing the new one. The exact steps vary according to the motherboard and case, but use the following general steps:

Power down the PC and all attached devices. Disconnect all external cables other than the power cord, noting which cable connects to which port. Then move the PC to your work area and remove the cover from the case. We can attest that one wayward case screw can destroy a vacuum cleaner, so put the screws safely aside. An old egg carton or ice cube tray makes a good parts organizer.
If the PC power cord is connected to an outlet strip, surge suppressor, or UPS, turn off the main power switch on that device, and turn off the main PC power switch as well. This removes power from the PC, but leaves the PC grounded.

With nearly all AT form factor power supplies and motherboards, turning off the PC power switch actually removes all power from the motherboard. With newer ATX (and variants) power supplies and motherboards, turning off the main PC power switch leaves some power flowing to the motherboard, which supports such features as Wake-on-Ring (WOR), Wake-on-LAN (WOL), and Suspend to RAM (STR).

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

When you upgrade an existing system without replacing the motherboard, the BIOS version it uses can be a critical issue. Some system features—notably support for large hard disks, high-speed transfer modes, and AGP—are BIOS-dependent, so an in-place upgrade often requires a BIOS upgrade as well. Fortunately, recent systems use a flash BIOS, which can be upgraded simply by downloading a later version of the BIOS to replace the existing BIOS.

Be extraordinarily careful when upgrading a flash BIOS. Before you proceed, make absolutely sure that the BIOS upgrade patch you are about to apply is the exact one required for the current BIOS. If you apply the wrong patch, you may render your system unbootable from the floppy drive, which makes it difficult or impossible to recover by reapplying the proper patch.

Upgrading a flash BIOS requires two files. The first is the upgraded BIOS itself in binary form. The second is the "flasher" program provided by the BIOS manufacturer, e.g., awdflash.exe. The exact steps you follow to upgrade a flash BIOS vary slightly according to the BIOS manufacturer and the version of the flasher program you are using, but the following steps are typical:

Before proceeding, record all current BIOS settings, either using pencil and paper or a utility program that writes BIOS settings to a disk file. If you have a UPS, connect the system to it for the duration of the BIOS update. Losing power during a BIOS update can result in a motherboard that is unusable and must be returned to the maker for repair.
Determine the manufacturer, version, date, and identifying string of the existing BIOS. You can do this by using a third-party diagnostics program like CheckIt, or by watching the BIOS screen that appears briefly each time the system boots. With most systems, pressing the Pause key halts the boot screen, allowing you to record the BIOS information at your leisure. With other systems, the Pause key does nothing, so you may have to reboot the system several times to record all the relevant information. It is important to record

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

For our latest motherboard recommendations by brand name and model number, visit:

http://www.hardwareguys.com/picks/motherboards.html

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

The processor, also called the microprocessor or CPU (for Central Processing Unit), is the brain of the PC. It performs all general computing tasks and coordinates tasks done by memory, video, disk storage, and other system components. The CPU is a very complex chip that resides directly on the motherboard of most PCs, but may sometimes reside on a daughtercard that connects to the motherboard via a dedicated specialized slot.

A processor executes programs—including the operating system itself and the user applications—all of which perform useful work. From the processor's point of view, a program is simply a group of low-level instructions that it executes more or less in sequence as it receives them. How efficiently and effectively the processor executes instructions is determined by its internal design, also called its architecture. The CPU architecture, in conjunction with CPU speed, determines how fast the CPU executes instructions of various types. The external design of the processor, specifically its external interfaces, determines how fast it communicates information back and forth with external cache, main memory, the chipset, and other system components.

Modern processors have the following internal components:

Execution unit: The core of the CPU, the execution unit processes instructions.
Branch predictor: The branch predictor attempts to guess where the program will jump (branch) next, allowing the Prefetch and Decode Unit to retrieve instructions and data in advance so they will already be available when the CPU requests them.
Floating-point unit

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

A processor executes programs—including the operating system itself and the user applications—all of which perform useful work. From the processor's point of view, a program is simply a group of low-level instructions that it executes more or less in sequence as it receives them. How efficiently and effectively the processor executes instructions is determined by its internal design, also called its architecture. The CPU architecture, in conjunction with CPU speed, determines how fast the CPU executes instructions of various types. The external design of the processor, specifically its external interfaces, determines how fast it communicates information back and forth with external cache, main memory, the chipset, and other system components.

Modern processors have the following internal components:

Execution unit: The core of the CPU, the execution unit processes instructions.
Branch predictor: The branch predictor attempts to guess where the program will jump (branch) next, allowing the Prefetch and Decode Unit to retrieve instructions and data in advance so they will already be available when the CPU requests them.
Floating-point unit: The floating-point unit (FPU) is a specialized logic unit optimized to perform non-integer calculations much faster than the general-purpose logic unit can perform them.
Primary cache: Also called Level 1or L1 cache, primary cache is a small amount of very fast memory that allows the CPU to retrieve data immediately, rather than waiting for slower main memory to respond. See Chapter 5.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Nearly all PCs use either an Intel CPU or an Intel-compatible CPU made by AMD (K6/Athlon/Duron series). The dominance of Intel in CPUs and Microsoft in operating systems gave rise to the hybrid term Wintel, which refers to systems that run Windows on an Intel or compatible CPU. Intel processors are referred to generically as x86 processors, based on Intel's early processor naming convention, 8086, 80186, 80286, etc. Intel has produced seven CPU generations, the first five of which are now obsolete and the sixth obsolescent.

First generation: The 8086 was Intel's first mainstream processor, and used 16 bits for both internal and external communications. The 8086 was first used in the late 1970s in dedicated word processors and minicomputers like the DisplayWriter and the System/23 DataMaster. When IBM shipped their first PC in 1981, they used the 8088, an 8086 variant that used 16 bits internally but only 8 bits externally, because 8-bit peripherals were at that time more readily available and less expensive than were 16-bit components. The 8086 achieved prominence much later when Compaq created the DeskPro as an improved clone of the IBM PC/XT. A few early PCs, notably Radio Shack models, were also built around the 80186 and 80188 CPUs, which were enhanced versions of the 8086 and 8088 respectively. The 8088 and 8086 CPUs did not include a floating-point unit (FPU), although an 8087 FPU, called a math coprocessor, was available as an optional upgrade chip. First-generation Intel CPUs (or their modern equivalents) are still used in some embedded applications, but they are long obsolete as general-purpose CPUs.
Second generation: In 1982, Intel introduced the long-awaited follow-on to their first-generation processors. The 80286, based on the iAPX-32 core, provided a quantum leap in processor performance, executing instructions as much as five times faster than an 808x processor running at the same clock speed. The 80286 processed instructions as fast as many mainframe processors of the time. The 80286 also increased addressable memory from 1 MB to 16 MB, and introduced

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Until late 1999, Intel had the desktop processor market largely to itself. There were competing incompatible systems such as the Apple Mac, based on processors from Motorola, IBM, and others, but those systems sold in relatively small numbers. Some companies, including Cyrix, IDT, Harris, and AMD itself, made Intel-compatible processors, but those were invariably a step behind Intel's flagship processors. When those companies—which Intel calls "imitators"—were producing enhanced 286s, Intel was already shipping the 386 in volume. When the imitators began producing enhanced 386-compatible processors, Intel had already begun shipping the 486, and so on. Each time Cyrix, AMD, and the others got a step up, Intel would turn around and release the next-generation processor. As a result, these other companies' processors sold at low prices and were used largely in low-end systems. No one could compete with Intel in their core market.

All of that changed dramatically in late 1999, when AMD began shipping a processor called the Athlon. The Athlon didn't just match the best Intel processors. It was faster than the best Intel could produce, and was in many respects a more sophisticated processor. Intel had a fight on its hands, and it's remained so to the present day.

If you ever take a moment to appreciate how much processor you can get for so little money nowadays, give thanks to AMD. Without AMD, we'd all still be running sixth-generation Intel processors at 500 MHz or so. An entry-level Intel processor would cost $200 or $250, and a high-end one (that might run at 750 MHz) would probably cost $1,000 or more. The presence of AMD as a worthy competitor meant that Intel could no longer play the game of releasing faster processors in dribs and drabs at very high prices. Instead, they had to fight for their lives by shipping faster and faster processors at lower and lower prices. We all have AMD to thank for that, and Intel should thank AMD as well. Although we're sure Intel wishes AMD would just disappear (and vice versa), the fact is that the competition has made both Intel and AMD better companies, as well as providing the obvious benefits to us, the users.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

The processor you choose determines how fast the system runs, and how long it will provide subjectively adequate performance before you need to replace the processor or the system itself. Buying a processor just fast enough to meet current needs means that you'll have to upgrade in a few months. But processor pricing has a built-in law of diminishing returns. Spending twice as much on a processor doesn't buy twice the performance. In fact, you'll be lucky to get 25% more performance for twice the money. So although it's a mistake to buy too slow a processor, it's also a mistake to buy one that's too fast. Consider the following issues when choosing a processor:

Horizon: What kind of applications do you run and how long do you want the system to be usable without requiring an upgrade? If you run mostly standard productivity applications and don't upgrade them frequently, a low- to low/midrange Celeron or Duron may still be fast enough a year or more after you buy it. If you run cutting-edge games or other demanding applications, buy a midrange or faster processor initially, and expect to replace it every six months to a year. But expect to pay a price for remaining on the bleeding edge.
Hassle: Do you mind upgrading your system frequently? If you don't mind replacing the processor every year, you can get most of the performance of a high-end system at minimal cost by replacing the processor every year or so with the then-current midrange processor.
Trade-offs: If you're working on a fixed budget, don't spend too much on the processor to the detriment of the rest of the system. You may be better off, for example, settling for a midrange Celeron and spending the extra money on more memory, a faster hard disk, or a bigger monitor. A Celeron/1200 with 256 MB RAM blows the doors off a Pentium 4/2.0G with 32 MB RAM every day of the week. Don't make yourself "processor poor."

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

The following sections describe the steps required to install and configure standard slotted and socketed processors. The steps we describe are generally applicable to any modern processor of a given type, but the details may vary slightly between different processors, particularly with regard to such things as configuring the motherboard and installing heatsink/fan units. If this is the first time you've installed a processor, or if you are in doubt about any step, refer to the documentation provided by the manufacturer of your specific processor and motherboard.

Before you install any processor, make sure that you have identified exactly both the processor itself and the motherboard you plan to install it in. If the processor is not new, you can identify it using the steps described earlier in this chapter. All high-quality motherboards have information printed on the board itself that identifies the manufacturer, model, and revision number. If the board does not contain such information, you may be able to identify the board by writing down the full BIOS string displayed by the BIOS boot screen and checking that string against one of the BIOS sites listed in Chapter 3. However, such "anonymous" boards are usually of very low quality, so it's generally better to replace such a board rather than attempt to use it.

Before you install a processor, make absolutely sure the processor is compatible with the motherboard. It is not safe to assume that merely because the processor fits the socket or slot that it will function properly in that motherboard. In some cases, the processor simply will not work. For example, there are many incompatibilities between Socket 370 processors and motherboards. Not all Socket 370 processors can be used in all Socket 370 motherboards, even if processor and motherboard were both made by Intel. In that situation, no damage is done. The processor simply doesn't work. There are, however, two common situations in which installing an incorrect processor may damage the processor and/or the motherboard:

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

Although processor makers probably hate us for saying so, the processor actually plays a relatively minor role in overall system performance. The difference in absolute processor performance between a $50 processor and a $500 processor may be a factor of two or less. Nor does buying a $500 processor make your system run twice as fast, because processor speed is only one element of system performance. Before you plunk down $500 for a processor, consider instead spending that extra money on more memory, a faster video card, a SCSI hard drive, or all of those.

Inexpensive systems (sub $1,000): Intel Celeron 4. We recommend the Intel Celeron 4 processor largely because of the many top-notch integrated motherboards available for it. Yes, you can find cheap Pentium 4 and Athlon systems in this price range, but those systems are built to a price point using components of lower quality than we feel comfortable using. The AMD Duron is also an attractive processor. Many inexpensive systems are built around the Duron, which has higher performance clock-for-clock than the Celeron, but most inexpensive Duron systems use VIA-based motherboards, which we find unacceptable. For a stable system with reasonably good performance, it's hard to beat the Celeron. We'd choose the fastest retail-boxed Celeron you can find for about $85.
Mainstream and performance systems ($1,000 to $2,500+): Intel Pentium 4. For mainstream and performance systems, we recommend the Intel Pentium 4. Pentium 4 processors do not match the AMD Athlon in performance clock-for-clock, and cost more for equivalent performance, but once again the superiority of Intel chipsets and motherboards in terms of stability, reliability, and build quality is more than enough to offset the price and performance advantages of AMD Athlon processors. As with the Celerons, the fastest Pentium 4 models sell at a high premium compared to models that are only slightly slower. We'd choose the fastest Socket 478 Northwood-core Pentium 4 processor available for $175 or so.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

A few years ago, this book could have covered memory in a page or two. All memory came as discrete chips, and there were only a few sizes and speeds available. You bought however many chips you needed, installed them, and that was that.

Dramatic CPU speed increases over the last few years have brought designers up against one hard fact. It's difficult to design faster CPUs, but it's even harder to build faster memory. Building affordable fast memory is harder still. Faster processors require faster memory, so engineers have come up with various methods to increase memory speeds. One result is that memory isn't a simple issue any more. This chapter tells you what you need to know to make good decisions about buying memory when you build or upgrade a system.

This chapter focuses on general-purpose memory, where PCs store programs and data that are currently in use, the pipeline that supplies data to and receives results from the processor. General-purpose memory, called read-write memory or random access memory (RAM), must be readable-from and writable-to. Two types of RAM are used on modern PCs:

Dynamic RAM (DRAM): Dynamic RAM stores data for only a tiny fraction of a second before losing it. To maintain stored data, the system must constantly refresh DRAM, which exacts a performance penalty and limits its speed. Typical DRAM provides 60 ns access, but costs only a dollar per megabyte.
Static RAM (SRAM): Static RAM automatically maintains its contents as long as power is applied to it, without requiring refresh. SRAM is the fastest memory type available, but is very expensive and power-hungry.

PCs use a tiered memory architecture that takes advantage of these characteristics:

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy reprint rights for this chapter

This chapter focuses on general-purpose memory, where PCs store programs and data that are currently in use, the pipeline that supplies data to and receives results from the processor. General-purpose memory, called read-write memory or random access memory (RAM), must be readable-from and writable-to. Two types of RAM are used on modern PCs:

Dynamic RAM (DRAM): Dynamic RAM stores data for only a tiny fraction of a second before losing it. To maintain stored data, the system must constantly refresh DRAM, which exacts a performance penalty and limits its speed. Typical DRAM provides 60 ns access, but costs only a dollar per megabyte.
Static RAM (SRAM): Static RAM automatically maintains its contents as long as power is applied to it, without requiring refresh. SRAM is the fastest memory type available, but is very expensive and power-hungry.

PCs use a tiered memory architecture that takes advantage of these characteristics:

Main memory: The bulk of a PC's memory uses DRAM and is called main memory . It is large—typically 32 MB to 256 MB or more—but too slow to keep up with a modern CPU. Main memory is where the CPU stores programs and data that it will soon need. Main memory functions as a buffer between the CPU and disk and stores tens to hundreds of megabytes.
Cache memory: Cache memory is a small amount of fast SRAM that buffers access between the CPU and main memory.

Modern PCs have at least one, typically two, and sometimes three layers of cache memory:

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·

Table of Contents