Advanced Technology Extended

Intel developed the Advanced Technology eXtended (ATX) motherboard in the mid-1990s to improve upon the classic AT-style motherboard architecture that had ruled the PC world for many years. The ATX motherboard has the processor and memory slots at right angles to the expansion cards, like the one in Figure 1.1. This arrangement puts the processor and memory in line with the fan output of the power supply, allowing the processor to run cooler. And because those components are not in line with the expansion cards, you can install full-length expansion cards—adapters that extend the full length of the inside of a standard computer case—in an ATX motherboard machine. ATX (and its derivatives, such as micro-ATX) is the primary PC motherboard form factor in use today. Standard ATX motherboards measure 12″ × 9.6″ (305 mm × 244 mm).

Information Technology eXtended

The Information Technology eXtended (ITX) line of motherboard form factors was developed by VIA Technologies in the early 2000s as a low-power, small form factor (SFF) board for specialty uses, including home-theater systems, compact desktop systems, gaming systems, and embedded components. ITX itself is not an actual form factor but a family of form factors. The family consists of the following form factors:

• Mini-ITX—6.7″ × 6.7″ (170 mm × 170 mm)
• Nano-ITX—4.7″ × 4.7″ (120 mm × 120 mm)
• Pico-ITX—3.9″ × 2.8″ (100 mm × 72 mm)
• Mobile-ITX—2.4″ × 2.4″ (60 mm × 60 mm)

The mini-ITX motherboard has four mounting holes that line up with three or four of the holes in the ATX and micro-ATX form factors. In mini-ITX boards, the rear interfaces are placed in the same location as those on the ATX motherboards. These features make mini-ITX boards compatible with ATX cases. This is where the mounting compatibility ends, because despite the PC compatibility of the other ITX form factors, they are used in embedded systems, such as set-top boxes, home entertainment systems, and smartphones, and lack the requisite mounting and interface specifications. Figure 1.2 shows the three larger forms of ITX motherboards, next to two ATX motherboards for comparison.

Bus Architecture

In a PC, data is sent from one component to another via a bus, which is a common collection of signal pathways. In the very early days, PCs used serial buses, which sent one bit at a time and were painfully slow. Brilliant engineers realized that they could redesign the bus and send 8 bits at a time (over synchronized separate lines), which resulted in a big speed increase. This was known as a parallel bus.

The downside of parallel communications is the loss of circuit length (how long the circuit could be) and throughput (how much data could move at one time). The signal could travel only a short distance, and the amount of data was limited due to the careful synchronization needed between separate lines, the speed of which must be controlled to limit skewing the arrival of the individual signals at the receiving end.

What was once old is new again, as engineers have discovered methods to make serial transmissions work at data rates that are many times faster than parallel signals. Therefore, nearly everything you see today uses a serial bus. The only limitation of serial circuits is in the capability of the transceivers, which tends to grow over time at a refreshing rate due to technical advancements. Examples of specifications that have heralded the dominance of serial communications are Serial Advanced Technology Attachment (Serial ATA, or SATA), Universal Serial Bus (USB), IEEE 1394/FireWire, and Peripheral Component Interconnect Express (PCIe).

On a motherboard, several different buses are used. Expansion slots of various architectures, such as PCIe, are included to allow for the insertion of external devices or adapters. Other types of buses exist within the system to allow communication between the CPU, RAM, and other components with which data must be exchanged. Except for CPU slots and sockets and memory slots, there are no insertion points for devices in many closed bus systems because no adapters exist for such an environment.

The various buses throughout a given computer system can be rated by their bus speeds. The higher the bus speed, the higher its performance. In some cases, various buses must be synchronized for proper performance, such as the system bus and any expansion buses that run at the front-side bus speed. Other times, one bus will reference another for its own speed. The internal bus speed of a CPU is derived from the front-side bus clock, for instance. The buses presented throughout this chapter are accompanied by their speeds, where appropriate.

Chipsets

A chipset is a collection of chips or circuits that perform interface and peripheral functions for the processor. This collection of chips is usually the circuitry that provides interfaces for memory, expansion cards, and onboard peripherals, and it generally dictates how a motherboard will communicate with the installed peripherals.

Chipsets are usually given a name and model number by the original manufacturer. For example, B550 and X570 are chipsets that support Advanced Micro Devices, Inc. (AMD) processors, and Z490 and H410 are Intel motherboard chipsets. Typically, the manufacturer and model tell you that your particular chipset has a certain set of features (for example, type of CPU and RAM supported, type and brand of onboard video, and so on). Don't worry about memorizing any chipset names—you can look them up online to understand their features.

Chipsets can be made up of one or several integrated circuit chips. Intel-based motherboards, for example, typically use two chips. To know for sure, you must check the manufacturer's documentation, especially because cooling mechanisms frequently obscure today's chipset chips, sometimes hindering visual brand and model identification.

Chipsets can be divided into two major functional groups, called Northbridge and Southbridge. Let's take a brief look at these groups and the functions they perform.

Southbridge

The Southbridge subset of the chipset is responsible for providing support to the slower onboard peripherals (USB, Serial and Parallel ATA, parallel ports, serial ports, and so on), managing their communications with the rest of the computer and the resources given to them. These components do not need to keep up with the external clock of the CPU and do not represent a bottleneck in the overall performance of the system. Any component that would impose such a restriction on the system should eventually be developed for FSB attachment.

In other words, if you're considering any component other than the CPU, memory and cache, or PCIe slots, the Southbridge is in charge. Most motherboards today have integrated USB, network, and analog and digital audio ports for the Southbridge to manage, for example, all of which are discussed in more detail later in this chapter or in Chapter 3, “Peripherals, Cables, and Connectors.” The Southbridge is also responsible for managing communications with the slower expansion buses, such as PCI, and legacy buses.

Figure 1.3 is a photo of the chipset of a motherboard, with the heat sink of the Northbridge at the top left, connected to the heat-spreading cover of the Southbridge at the bottom right.

Figure 1.4 shows a schematic of a typical motherboard chipset (both Northbridge and Southbridge) and the components with which they interface. Notice which components interface with which parts of the chipset.

Expansion Slots

The most visible parts of any motherboard are the expansion slots. These are small plastic slots, usually from 1 to 6 inches long and approximately ½ inch wide. As their name suggests, these slots are used to install various devices in the computer to expand its capabilities. Some expansion devices that might be installed in these slots include video, network, sound, and disk interface cards.

If you look at the motherboard in your computer, you will more than likely see one of the main types of expansion slots used in computers today, which are PCI and PCIe. In the following sections, we will cover how to visually identify the different expansion slots on the motherboard.

PCIe Expansion Slots

The most common expansion slot architecture that is being used by motherboards is PCI Express (PCIe). It was designed to be a replacement for PCI, as well as an older video card standard called accelerated graphics port (AGP). PCIe has the advantage of being faster than AGP while maintaining the flexibility of PCI. PCIe has no plug compatibility with either AGP or PCI. Some modern PCIe motherboards can be found with regular PCI slots for backward compatibility, but AGP slots have not been included for many years.

PCIe is casually referred to as a bus architecture to simplify its comparison with other bus technologies. True expansion buses share total bandwidth among all slots, each of which taps into different points along the common bus lines. In contrast, PCIe uses a switching component with point-to-point connections to slots, giving each component full use of the corresponding bandwidth and producing more of a star topology versus a bus. Furthermore, unlike other expansion buses, which have parallel architectures, PCIe is a serial technology, striping data packets across multiple serial paths to achieve higher data rates.

PCIe uses the concept of lanes, which are the switched point-to-point signal paths between any two PCIe components. Each lane that the switch interconnects between any two intercommunicating devices comprises a separate pair of wires for both directions of traffic. Each PCIe pairing between cards requires a negotiation for the highest mutually supported number of lanes. The single lane or combined collection of lanes that the switch interconnects between devices is referred to as a link.

There are seven different link widths supported by PCIe, designated x1 (pronounced “by 1”), x2, x4, x8, x12, x16, and x32, with x1, x4, and x16 being the most common. The x8 link width is less common than these but more common than the others. A slot that supports a particular link width is of a physical size related to that width because the width is based on the number of lanes supported, requiring a related number of wires. As a result, an x8 slot is longer than an x1 slot but shorter than an x16 slot. Every PCIe slot has a 22-pin portion in common toward the rear of the motherboard, which you can see in Figure 1.7, in which the rear of the motherboard is to the left. These 22 pins comprise mostly voltage and ground leads. (The PCIe slots are the longer and lighter ones in Figure 1.6.)

Four major versions of PCIe are currently available in the market: 1.x, 2.x, 3.0, and 4.0. For the four versions, a single lane, and therefore an x1 slot, operates in each direction (or transmits and receives from either communicating device's perspective), at a data rate of 250 MBps (almost twice the rate of the most common PCI slot), 500 MBps, approximately 1 GBps, and roughly 2 GBps, respectively.

An associated bidirectional link has a nominal throughput of double these rates. Use the doubled rate when comparing PCIe to other expansion buses because those other rates are for bidirectional communication. This means that the 500 MBps bidirectional link of an x1 slot in the first version of PCIe was comparable to PCI's best, a 64-bit slot running at 66 MHz and producing a throughput of 533 MBps.

Combining lanes simply results in a linear multiplication of these rates. For example, a PCIe 1.1 x16 slot is capable of 4 GBps of throughput in each direction, 16 times the 250 MBps x1 rate. Bidirectionally, this fairly common slot produces a throughput of 8 GBps. Each subsequent PCIe specification doubles this throughput. The aforementioned PCIe 5.0 will produce bidirectional throughput of approximately 128 GBps, which is faster than some DDR4 standards (which is to say, it's really, really fast).

Because of its high data rate, PCIe is the current choice of gaming aficionados. Additionally, technologies similar to NVIDIA's Scalable Link Interface (SLI) allow such users to combine preferably identical graphics adapters in appropriately spaced PCIe x16 slots with a hardware bridge to form a single virtual graphics adapter. The job of the bridge is to provide non-chipset communication among the adapters. The bridge is not a requirement for SLI to work, but performance suffers without it. SLI-ready motherboards allow two, three, or four PCIe graphics adapters to pool their graphics processing units (GPUs) and memory to feed graphics output to a single monitor attached to the adapter acting as the primary SLI device. SLI implementation results in increased graphics performance over single-PCIe and non-PCIe implementations.

Refer back to Figure 1.6, which is a photo of an SLI-ready motherboard with three PCIe x16 slots (every other slot, starting with the top one), one PCIe x1 slot (second slot from the top), and two PCI slots (first and third slots from the bottom). Notice the latch and tab that secures the x16 adapters in place by their hooks. Any movement of these high-performance devices can result in temporary failure or poor performance.

Memory Slots and Cache

Memory, or random access memory (RAM), slots are the next most notable slots on a motherboard. These slots are designed for the modules that hold memory chips that make up primary memory, which is used to store currently used data and instructions for the CPU. Many types of memory are available for PCs today. In this chapter, you will become familiar with the appearance and specifications of the slots on the motherboard so that you can identify them and appropriately install or replace RAM.

For the most part, PCs today use memory chips arranged on a small circuit board. A dual in-line memory module (DIMM) is one type of circuit board. Today's DIMMs differ in the number of conductors, or pins, that each particular physical form factor uses. Some common examples include 168-, 184-, 240-, and 288-pin configurations. In addition, laptop memory comes in smaller form factors known as small outline DIMMs (SODIMMs) and MicroDIMMs. More detail on memory packaging and the technologies that use them can be found later in this chapter in the section “Understanding Memory.”

Memory slots are easy to identify on a motherboard. Classic DIMM slots were usually black and, like all memory slots, were placed very close together. DIMM slots with color-coding are more common these days, however. The color-coding of the slots acts as a guide to the installer of the memory. See the section “Single-, Dual-, Triple-, and Quad-Channel Memory” later in this chapter for more on the purpose of this color-coding. Consult the motherboard's documentation to determine the specific modules allowed as well as their required orientation. The number of memory slots varies from motherboard to motherboard, but the structure of the different slots is similar. Metal pins in the bottom make contact with the metallic pins on each memory module. Small metal or plastic tabs on each side of the slot keep the memory module securely in its slot. Figure 1.8 shows four memory slots, with the CPU socket included for reference.

Sometimes, the amount of primary memory installed is inadequate to service additional requests for memory resources from newly launched applications. When this condition occurs, the user may receive an “out of memory” error message and an application may fail to launch. One solution for this is to use the hard drive as additional RAM. This space on the hard drive is known as a swap file or a paging file. The technology in general is known as virtual memory or virtual RAM. The paging file is called PAGEFILE.SYS in modern Microsoft operating systems. It is an optimized space that can deliver information to RAM at the request of the memory controller faster than if it came from the general storage pool of the drive. It's located at c:\pagefile.sys by default. Note that virtual memory cannot be used directly from the hard drive; it must be paged into RAM as the oldest contents of RAM are paged out to the hard drive to make room. The memory controller, by the way, is the chip that manages access to RAM as well as adapters that have had a few hardware memory addresses reserved for their communication with the processor.

Nevertheless, relying too much on virtual memory (check your page fault statistics in the Reliability and Performance Monitor) results in the entire system slowing down noticeably. An inexpensive and highly effective solution is to add physical memory to the system, thus reducing its reliance on virtual memory. More information on virtual memory and its configuration can be found in Chapter 13, “Operating System Basics.”

Another type of memory common in PCs is cache memory, which is small and fast and logically sits between the CPU and RAM. Cache is a very fast form of memory forged from static RAM, which is discussed in detail in the section “Understanding Memory” later in this chapter. Cache improves system performance by predicting what the CPU will ask for next and prefetching this information before being asked. This paradigm allows the cache to be smaller in size than the RAM itself. Only the most recently used data and code or that which is expected to be used next is stored in cache.

You'll see three different cache designations:

• Level 1 Cache   L1 cache is the smallest and fastest, and it's on the processor die itself. In other words, it's an integrated part of the manufacturing pattern that's used to stamp the processor pathways into the silicon chip. You can't get any closer to the processor than that.

  Though the definition of L1 cache has not changed much over the years, the same is not true for other cache levels. L2 and L3 cache used to be on the motherboard but now have moved on-die in most processors as well. The biggest differences are the speed and whether they are shared.

• Level 2 Cache   L2 cache is larger but a little slower than L1 cache. For processors with multiple cores, each core will generally have its own dedicated L1 and L2 caches. A few processors share a common L2 cache among the cores.
• Level 3 Cache   L3 cache is larger and slower than L1 or L2, and is usually shared among all processor cores.

The typical increasing order of capacity and distance from the processor die is L1 cache, L2 cache, L3 cache, RAM, and HDD/SSD (hard disk drive and solid-state drive—more on these in Chapter 2). This is also the typical decreasing order of speed. The following list includes representative capacities of these memory types. The cache capacities are for each core of the 10th generation Intel Core i7 processor. The other capacities are simply modern examples.

• L1 cache—80 KB (32 KB for instructions and 48 KB for data)
• L2 cache—512 KB
• L3 cache—8–16 MB
• RAM—16–256 GB
• HDD/SSD—100s of GB to several TB

One way to find out how much cache your system has is to use a utility such as CPU-Z, as shown in Figure 1.9. CPU-Z is freeware that can show you the amount of cache, processor name and number, motherboard and chipset, and memory specifications. It can be found at www.cpuid.com.

Central Processing Unit and Processor Socket

The “brain” of any computer is the central processing unit (CPU). There's no computer without a CPU. There are many different types of processors for computers—so many, in fact, that you will learn about them later in this chapter in the section “Understanding Processors.”

Typically, in today's computers, the processor is the easiest component to identify on the motherboard. It is usually the component that has either a fan or a heat sink (usually both) attached to it (as shown in Figure 1.10). These devices are used to draw away and disperse the heat that a processor generates. This is done because heat is the enemy of microelectronics. Today's processors generate enough heat that, without heat dispersal, they would permanently damage themselves and the motherboard in a matter of minutes, if not seconds.

CPU sockets are almost as varied as the processors that they hold. Sockets are basically flat and have several columns and rows of holes or pins arranged in a square, as shown in Figure 1.11. The left socket is known as Socket AM4, made for AMD processors such as the Ryzen, and has holes to receive the pins on the CPU. This is known as a pin grid array (PGA) arrangement for a CPU socket. The holes and pins are in a row/column orientation, an array of pins. The right socket is known as LGA 1200, and there are spring-loaded pins in the socket and a grid of lands on the CPU. The land grid array (LGA) is a newer technology that places the delicate pins (yet more sturdy than those on chips) on the cheaper motherboard instead of on the more expensive CPU, opposite to the way that the aging PGA does. The device with the pins has to be replaced if the pins become too damaged to function. The PGA and LGA are mentioned again later in this chapter in the section “Understanding Processors.”

Modern CPU sockets have a mechanism in place that reduces the need to apply considerable force to the CPU to install a processor, which was necessary in the early days of personal computing. Given the extra surface area on today's processors, excessive pressure applied in the wrong manner could damage the CPU packaging, its pins, or the motherboard itself. For CPUs based on the PGA concept, zero insertion force (ZIF) sockets are exceedingly popular. ZIF sockets use a plastic or metal lever on one of the two lateral edges to lock or release the mechanism that secures the CPU's pins in the socket. The CPU rides on the mobile top portion of the socket, and the socket's contacts that mate with the CPU's pins are in the fixed bottom portion of the socket. The image of Socket AM4 shown on the left in Figure 1.11 illustrates the ZIF locking mechanism at the right edge of the socket.

For processors based on the LGA concept, a socket with a different locking mechanism is used. Because there are no receptacles in either the motherboard or the CPU, there is no opportunity for a locking mechanism that holds the component with the pins in place. LGA-compatible sockets, as they're called despite the misnomer, have a lid that closes over the CPU and is locked in place by an L-shaped arm that borders two of the socket's edges. The nonlocking leg of the arm has a bend in the middle that latches the lid closed when the other leg of the arm is secured. The right image in Figure 1.11 shows an LGA socket with no CPU installed and the locking arm secured over the lid's tab, along the bottom edge.

Listing out all the desktop PC socket types you might encounter would take a long time. Instead, we'll give you a sampling of some that you might see. The first thing you might notice is that sockets are made for Intel or AMD processors, but not both. Keep that compatibility in mind when replacing a motherboard or a processor. Make sure that the processor and motherboard were designed for each other (even within the Intel or AMD families); otherwise, they won't fit each other and won't work. Table 1.1 lists some common desktop socket/CPU relationships. Servers and laptops/tablets generally have different sockets altogether, although some CPU sockets will support processors designed for desktops or servers.

Multisocket and Server Motherboards

When it comes to motherboard compatibility, the two biggest things to keep in mind are the processor type and the case. If either of those are misaligned with what the motherboard supports, you're going to have problems.

Thus far, as we've talked about desktop motherboards and their CPU sockets, we have shown examples of boards that have just one socket. There are motherboards that have more than one CPU socket and conveniently, they are called multisocket (typically written as two words) motherboards. Figure 1.12 shows a two-socket motherboard made by GIGABYTE. The two CPU sockets are easily identifiable and note that each CPU socket has eight dedicated memory slots.

Trying to categorize server motherboards can be a bit challenging. Servers are expected to do a lot more work than the average PC, so it makes sense that servers need more powerful hardware. Servers can, and quite often do, make do with a single processor on a “normal” PC motherboard. At the same time, there are motherboards designed specifically for servers that support multiple processors (two and four sockets are common) and have expanded memory and networking capabilities as well. Further, while server motherboards are often ATX-sized, many server manufacturers create custom boards to fit inside their chassis. Regardless, multisocket and server motherboards will generally use the same CPU sockets that other motherboards use.

Mobile Motherboards

In small mobile devices, space is at a premium. Some manufacturers will use standard small-factor motherboards, but most create their own boards to fit inside specific cases. An example of an oddly shaped Dell laptop motherboard is shown in Figure 1.13. When replacing a laptop motherboard, you almost always need to use one from the exact same model, otherwise it won't fit inside the case.

Nearly all laptop processors are soldered onto the motherboard, so you don't have to worry about CPU socket compatibility. If the CPU dies, you replace the entire motherboard. We will cover laptop components more extensively in Chapter 9, “Laptop and Mobile Device Hardware.”

Power Connectors

In addition to these sockets and slots on the motherboard, a special connector (the 24-pin white block connector shown in Figure 1.14) allows the motherboard to be connected to the power supply to receive power. This connector is where the ATX power connector (mentioned in Chapter 2 in the section “Understanding Power Supplies”) plugs in.

Onboard Nonvolatile Storage Connectors

Nearly all users store data, and the most widely used data storage device is a hard drive. Hard drives are great because they store data even when the device is powered off, which explains why they are sometimes referred to as nonvolatile storage. There are multiple types of hard drives, and we'll get into them in more detail in Chapter 2. Of course, those drives need to connect to the motherboard, and that's what we'll cover here.

Integrated Drive Electronics/Parallel ATA

At one time, integrated drive electronics (IDE) drives were the most common type of hard drive found in computers. Though often thought of in relation to hard drives, IDE was much more than a hard drive interface; it was also a popular interface for many other drive types, including optical drives and tape drives. Today, we call it IDE Parallel ATA (PATA) and consider it to be a legacy technology. Figure 1.15 shows two PATA interfaces; you can see that one pin in the center is missing (as a key) to ensure that the cable gets attached properly. The industry now favors Serial ATA instead.

Serial ATA

Serial ATA (SATA) began as an enhancement to the original ATA specifications, also known as IDE and, today, PATA. Technology is proving that the orderly progression of data in a single-file path is superior to placing multiple bits of data in parallel and trying to synchronize their transmission to the point that all of the bits arrive simultaneously. In other words, if you can build faster transceivers, serial transmissions are simpler to adapt to the faster rates than are parallel transmissions.

The first version of SATA, known as SATA 1.5 Gbps (and also by the less-preferred terms SATA I and SATA 150), used an 8b/10b-encoding scheme that requires 2 non-data overhead bits for every 8 data bits. The result is a loss of 20 percent of the rated bandwidth. The silver lining, however, is that the math becomes quite easy. Normally, you have to divide by 8 to convert bits to bytes. With 8b/10b encoding, you divide by 10. Therefore, the 150 MBps throughput for which this version of SATA was nicknamed is easily derived as 1/10 of the 1.5 Gbps transfer rate. The original SATA specification also provided for hot swapping at the discretion of the motherboard and drive manufacturers.

Similar math works for SATA 3 Gbps, tagged as SATA II and SATA 300, and SATA 6 Gbps, which you might hear called SATA III or SATA 600. Note that each subsequent version doubles the throughput of the previous version. Figure 1.16 shows four SATA headers on a motherboard that will receive the data cable. Note that identifiers silkscreened onto motherboards often enumerate such headers. The resulting numbers are not related to the SATA version that the header supports. Instead, such numbers serve to differentiate headers from one another and to map to firmware identifiers, often visible within the BIOS configuration utility.

Another version of SATA that you will see is external SATA (eSATA). As you might expect based upon the name, this technology was developed for devices that reside outside of the case, not inside it. Many motherboards have an eSATA connector built in. If not, you can buy an expansion card that has eSATA ports and plugs into internal SATA connectors. Figure 1.17 shows an example of how the two ports are different. Finally, SATA and eSATA standards are compatible. In other words, SATA 6 Gbps equals eSATA 6 Gbps.

M.2

The most recent development in expansion connections is M.2 (pronounced “M dot 2”). So far it's primarily used for hard drives, but other types of devices, such as Wi-Fi, Bluetooth, Global Positioning System (GPS), and near-field communication (NFC) adapters are built for M.2 as well. We will cover M.2 in more depth in Chapter 2, thanks to how important it is to storage solutions.

It's important to call out that M.2 is a form factor, not a bus standard. The form factor supports existing SATA, USB, and PCIe buses. This means that if you hook up a SATA device to an M.2 slot (with the appropriate connector), the device speed will be regulated by SATA standards. Figure 1.18 shows two M.2 connectors on a motherboard.

Motherboard Headers

From the time of the very first personal computer, there has been a minimum expectation as to the buttons and LEDs that should be easily accessible to the user. At first, they generally appeared on the front of the case. In today's cases, buttons and LEDs have been added and placed on the top of the case or on a beveled edge between the top and the front. They have also been left on the front or have been used in a combination of these locations. These buttons and lights, as well as other external connectors, plug into the motherboard through a series of pins known as headers. Examples of items that are connected using a header include:

• Power button
• Power light
• Reset button
• Drive activity lights
• Audio jacks
• USB ports

Headers for different connections are often spread throughout different locations on the motherboard—finding the right one can sometimes be a frustrating treasure hunt. Other headers are grouped together. For example, most of the headers for the items on the front or top panel of the case are often co-located. The purpose for the header will be printed on the motherboard, and while that may tell you what should connect there, it often lacks detail in how it should be connected. The motherboard manufacturer's website is a good place to go if you need a detailed diagram or instructions. Figure 1.19 shows several headers on a motherboard. On the left is a USB header, then a system fan header in the center, and a block of front panel headers on the right, including the hard drive light, reset button, chassis intrusion detector, and power light.

USB Ports

So many temporarily attached devices feature USB connectivity, such as USB keys (flash drives) and cameras, that front-panel connectivity is a must. Finding your way to the back of the system unit for a brief connection is hardly worth the effort in some cases. For many years, motherboards have supplied one or more 10-position headers for internal connectivity of front-panel USB ports. Because this header size is popular for many connectors, only 9 positions tend to have pins protruding, while the 10th position acts as a key, showing up in different spots for each connector type to discourage the connection of the wrong cable. Figure 1.20 shows USB headers on a motherboard. The labels “USB56” and “USB78” indicate that one block serves ports 5 and 6, while the other serves ports 7 and 8, all of which are arbitrary, based on the manufacturer's numbering convention. In each, the upper left pin is “missing,” which is the key.

BIOS / UEFI and the POST Routine

Firmware is the name given to any software that is encoded in hardware, usually a read-only memory (ROM) chip, and it can be run without extra instructions from the operating system. Most computers, large printers, and devices with no operating system use firmware in some sense. The best example of firmware is a computer's Basic Input/Output System (BIOS), which is burned into a chip. Also, some expansion cards, such as SCSI cards and graphics adapters, use their own firmware utilities for setting up peripherals.

The BIOS chip, also referred to as the ROM BIOS chip, is one of the most important chips on the motherboard. This special memory chip contains the BIOS system software that boots the system and allows the operating system to interact with certain hardware in the computer in lieu of requiring a more complex device driver to do so. The BIOS chip is easily identified: If you have a brand-name computer, this chip might have on it the name of the manufacturer and usually the word BIOS. For clones, the chip usually has a sticker or printing on it from one of the major BIOS manufacturers (AMI, Phoenix, Award, Winbond, and others). On later motherboards, the BIOS might be difficult to identify or it might even be integrated into the Southbridge, but the functionality remains regardless of how it's implemented.

The successor to the BIOS is the Unified Extensible Firmware Interface (UEFI). The extensible features of the UEFI allow for the support of a vast array of systems and platforms by allowing the UEFI access to system resources for storage of additional modules that can be added at any time. In the following section, you'll see how a security feature known as Secure Boot would not be possible with the classic BIOS. It is the extensibility of the UEFI that makes such technology feasible.

Figure 1.21 gives you an idea of what a modern BIOS/UEFI chip might look like on a motherboard. Despite the 1998 copyright on the label, which refers only to the oldest code present on the chip, this particular chip can be found on motherboards produced as late as 2009. Notice also the Reset CMOS jumper at the lower left and its configuration silkscreen at the upper left. You might use this jumper to clear the CMOS memory, discussed shortly, when an unknown password, for example, is keeping you out of the BIOS/UEFI configuration utility. The jumper in the photo is in the clear position, not the normal operating position. System bootup is typically not possible in this state.

At a basic level, the BIOS/UEFI controls system boot options such as the sequence of drives from which it will look for operating system boot files. The boot sequence menu from a BIOS/UEFI is shown in Figure 1.22. Other interface configuration options will be available too, such as enabling or disabling integrated ports or an integrated video card. A popular option on corporate computers is to disable the USB ports, which can increase security and decrease the risk of contracting a virus.

Most BIOS/UEFI setup utilities have more to offer than a simple interface for making selections and saving the results. For example, these utilities often offer diagnostic routines that you can use to have the BIOS/UEFI analyze the state and quality of the same components that it inspects during bootup, but at a much deeper level.

Consider the scenario where a computer is making noise and overheating. You can use the BIOS/UEFI configuration utility to access built-in diagnostics to check the rotational speed of the motherboard fans. If the fans are running slower than expected, the noise could be related to the bearings of one or more fans, causing them to lose speed and, thus, cooling capacity.

There is often also a page within the utility that gives you access to such bits of information as current live readings of the temperature of the CPU and the ambient temperature of the interior of the system unit. On such a page, you can set the temperature at which the BIOS/UEFI sounds a warning tone and the temperature at which the BIOS/UEFI shuts down the system to protect it. You can also monitor the instantaneous fan speeds, bus speeds, and voltage levels of the CPU and other vital landmarks to make sure that they are all within acceptable ranges. You might also be able to set a lower fan speed threshold at which the system warns you. In many cases, some of these levels can be altered to achieve such phenomena as overclocking, which is using the BIOS/UEFI to set the system clock higher than what the CPU is rated for, or undervolting, which is lowering the voltage of the CPU and RAM, which reduces power consumption and heat production.

BIOS / UEFI Security and Encryption

The BIOS/UEFI has always played a role in system security. Since the early days of the personal computer, the BIOS allowed the setting of two passwords—the user (or boot) password and the supervisor/administrator, or access, password. The boot password is required to leave the initial power-on screens and begin the process of booting an operating system. The administrator password is required before entering the BIOS/UEFI configuration utility. It is always a good idea to set the administrator password, but the boot password should not be set on public systems that need to boot on their own, in case of an unforeseen power cycle.

In more recent years, the role of the BIOS/UEFI in system security has grown substantially. BIOS/UEFI security has been extended to a point where the operating system is ready to take it over. The BIOS/UEFI was a perfect candidate to supervise security and integrity in a platform-independent way. Coupled with the Trusted Platform Module (TPM), a dedicated security coprocessor, or cryptoprocessor, the BIOS can be configured to boot the system only after authenticating the boot device. This authentication confirms that the hardware being booted to has been tied to the system containing the BIOS/UEFI and TPM, a process known as sealing. Sealing the devices to the system also prohibits the devices from being used after removing them from the system. For further security, the keys created can be combined with a PIN or password that unlocks their use or with a USB flash drive that must be inserted before booting.

Microsoft's BitLocker uses the TPM to encrypt the entire drive. Normally, only user data can be encrypted, but BitLocker encrypts operating-system files, the Registry, the hibernation file, and so on, in addition to those files and folders that file-level encryption secures. If any changes have occurred to the Windows installation, the TPM does not release the keys required to decrypt and boot to the secured volume. TPM is configured in Windows under Start ➢ Settings ➢ Update & Security ➢ Windows Security ➢ Device security, as shown in Figure 1.23.

Most motherboards come with a TPM chip installed, but if they don't, it's not possible to add one. In those situations, you can enable the same functionality by using a hardware security module (HSM). An HSM is a security device that can manage, create, and securely store encryption keys—it enables users to safely encrypt and decrypt data. An HSM can take a few different forms. The simplest is a USB or PCIe device that plugs into a system. It could be set up for file encryption and decryption, required for the computer to boot, or both. For large-scale solutions, HSM-enabled servers can provide crypto services to an entire network.

When a certain level of UEFI is used, the system firmware can also check digital signatures for each boot file it uses to confirm that it is the approved version and has not been tampered with. This technology is known as Secure Boot. An example of a BIOS/UEFI's boot security screen is shown in Figure 1.24. The boot files checked include option ROMs (defined in the following section), the boot loader, and other operating-system boot files. Only if the signatures are valid will the firmware load and execute the associated software.

The problem can now arise that a particular operating system might not be supported by the database of known-good signatures stored in the firmware. In such a situation, the system manufacturer can supply an extension that the UEFI can use to support that operating system—a task not possible with traditional BIOS-based firmware.

Some BIOS firmware can monitor the status of a contact on the motherboard for intrusion detection. If the feature in the BIOS is enabled and the sensor on the chassis is connected to the contact on the motherboard, the removal of the cover will be detected and logged by the BIOS. This can occur even if the system is off, thanks to the CMOS battery. At the next bootup, the BIOS will notify you of the intrusion. No notification occurs over subsequent boots unless additional intrusion is detected.

POST

A major function of the BIOS/UEFI is to perform a process known as a power-on self-test (POST). POST is a series of system checks performed by the system BIOS/UEFI and other high-end components, such as the SCSI BIOS and the video BIOS, known collectively as option ROMs. Among other things, the POST routine verifies the integrity of the BIOS/UEFI itself. It also verifies and confirms the size of primary memory. During POST, the BIOS also analyzes and catalogs other forms of hardware, such as buses and boot devices, as well as manages the passing of control to the specialized BIOS/UEFI routines mentioned earlier. The BIOS/UEFI is responsible for offering the user a key sequence to enter the configuration routine as POST is beginning. Finally, once POST has completed successfully, the BIOS/UEFI selects the boot device highest in the configured boot order and executes the master boot record (MBR) or similar construct on that device so that the MBR can call its associated operating system's boot loader and continue booting up.

The POST process can end with a beep code or displayed code that indicates the issue discovered. Each BIOS/UEFI publisher has its own series of codes that can be generated. Figure 1.25 shows a simplified POST display during the initial boot sequence of a computer.

CMOS and CMOS Battery

Your PC has to keep certain settings when it's turned off and its power cord is unplugged:

• Date
• Time
• Hard drive/optical drive configuration
• Memory
• CPU settings, such as overclocking
• Integrated ports (settings as well as enable/disable)
• Boot sequence
• Power management
• Virtualization support
• Security (passwords, Trusted Platform Module settings, LoJack)

Consider a situation where you added a new graphics adapter to your desktop computer, but the built-in display port continues to remain active, prohibiting the new interface from working. The solution might be to alter your BIOS/UEFI configuration to disable the internal graphics adapter, so that the new one will take over. Similar reconfiguration of your BIOS/UEFI settings might be necessary when overclocking—or changing the system clock speed—is desired, or when you want to set BIOS/UEFI-based passwords or establish TPM-based whole-drive encryption, as with Microsoft's BitLocker. While not so much utilized today, the system date and time can be altered in the BIOS/UEFI configuration utility of your system; once, in the early days of personal computing, the date and time actually might have needed to be changed this way.

Your PC keeps these settings in a special memory chip called the complementary metal oxide semiconductor (CMOS) memory chip. Actually, CMOS (usually pronounced see-moss) is a manufacturing technology for integrated circuits. The first commonly used chip made from CMOS technology was a type of memory chip, the memory for the BIOS/UEFI. As a result, the term CMOS stuck and is the accepted name for this memory chip.

The BIOS/UEFI starts with its own default information and then reads information from the CMOS, such as which hard drive types are configured for this computer to use, which drive(s) it should search for boot sectors, and so on. Any overlapping information read from the CMOS overrides the default information from the BIOS/UEFI. A lack of corresponding information in the CMOS does not delete information that the BIOS knows natively. This process is a merge, not a write-over. CMOS memory is usually not upgradable in terms of its capacity and might be integrated into the BIOS/UEFI chip or the Southbridge.

To keep its settings, integrated circuit-based memory must have power constantly. When you shut off a computer, anything that is left in this type of memory is lost forever. The CMOS manufacturing technology produces chips with very low power requirements. As a result, today's electronic circuitry is more susceptible to damage from electrostatic discharge (ESD). Another ramification is that it doesn't take much of a power source to keep CMOS chips from losing their contents.

To prevent CMOS from losing its rather important information, motherboard manufacturers include a small battery called the CMOS battery to power the CMOS memory, shown in the bottom-left corner of Figure 1.26. The batteries come in different shapes and sizes, but they all perform the same function. Most CMOS batteries look like large watch batteries or small cylindrical batteries. Today's CMOS batteries are most often of a long-life, non-rechargeable lithium chemistry.

x64/x86

CISC (pronounced like disk, but with a “c”) and RISC (pronounced risk) are examples of an instruction set architecture (ISA). Essentially, it's the set of commands that the processor can execute. Both types of chips, when combined with software, can ultimately perform all the same tasks. They just go about it differently. When programmers develop code, they develop it for a CISC or a RISC platform.

As the CISC name implies, instructions sent to the computer are relatively complex (as compared to RISC), and as such they can do multiple mathematical tasks with one instruction, and each instruction can take several clock cycles to complete. We'll talk more about CPU speeds in the “CPU Characteristics” section later, but for now, know that if a CPU is advertised as having 3.8 GHz speed, that means it can complete roughly 3.8 billion cycles in one second. The core of a processor can only do one thing at a time—it just does them very, very quickly so it looks like it's multitasking.

CISC was the original ISA for microprocessors, and the most well-known example of CISC technology is the x64/x86 platform popularized by Intel. AMD processors are CISC chips as well. So where did the terms x64 and x86 come from? First, just a bit more theory.

There is a set of data lines between the CPU and the primary memory of the system—remember the bus? The most common bus today is 64 bits wide, although there are still some 32-bit buses kicking around out there. In older days, buses could theoretically be as narrow as 2 bits, and 8-bit and 16-bit buses were popular for CPUs for several years. The wider the bus, the more data that can be processed per unit of time, and hence, more work can be performed. Internal registers in the CPU might be only 32 bits wide, but with a 64-bit system bus, two separate pipelines can receive information simultaneously. For true 64-bit CPUs, which have 64-bit internal registers and can run x64 versions of Microsoft operating systems, the external system data bus will always be 64 bits wide or some larger multiple thereof.

In the last paragraph we snuck in the term x64, and by doing that we also defined it. It refers to processors that are designed to work with 64 bits of data at a time. To go along with it, the operating system must also be designed to work with x64 chips.

Contrast that with processors that can handle only 32 bits of information at once. Those are referred to as x86 processors. You might look at that last sentence and be certain that we made a typo, but we didn't. For a long time when 32-bit processors were the fastest on the PC market, Intel was the dominant player. Their CPUs had names like 80386 (aka i386) and 80486 (i486) and were based on the older 16-bit 80286 and 8086. Since the i386 and i486 were the most popular standards, the term x86 sprung up to mean a 32-bit architecture. So even though it may seem counterintuitive due to the numbers, x64 is newer and faster than x86.

Advanced RISC Machine

Moving into the RISC architecture, the primary type of processor used today is known as an Advanced RISC Machine (ARM) CPU. Depending on who you talk to and which sources you prefer, there are conflicting stories on if that's actually the right acronym, as ARM is also known as an Acorn RISC Machine. Regardless of what it stands for, ARM is a competing technology to Intel and AMD x64-based CPUs.

Based on the RISC acronym, one might think that the reduced set of instructions the processor can perform makes it inferior somehow, but that's not the case. Tasks just need to get executed in different ways. To use a human example, let's say that we tell you to add the number 7 to itself seven times. One way to do that is to use one step of multiplication: 7 × 7 equals 49. But what if we said you can't use multiplication? You can still get to the answer by using addition. It will just take you seven steps instead of one. Same answer, different process. That's kind of how RISC compares to CISC.

RISC processors have some advantages over their CISC counterparts. They can be made smaller than CISC chips and they produce less heat, making them ideal for mobile devices. In fact, nearly all smartphones use RISC-based chips, such as Apple's A15 and Samsung's Exynos series processors. On the downside, RISC processors use more memory than CISC ones do because it takes more code to complete a task with a RISC chip.

Like x64/x86, ARM processors have evolved over time—64-bit implementations are the most current, and they are designated ARM64; 32-bit versions are known simply as ARM.

CPU Cores

Older processors were single-core, meaning that there was one set of instruction pathways through the processor. Therefore, they could process one set of tasks at a time. Designers then figured out how to speed up the computer by creating multiple cores within one processor package. Each core effectively operates as its own independent processor, provided that the operating system and applications are able to support multicore technology. (Nearly all do.)

Today, almost all desktop CPUs in the market are multicore. The number of cores you want may determine the processor to get. For example, the 10th-generation Intel Core i7 has eight cores whereas the i5 has six.

Speed

The speed of the processor is generally described in clock frequency. Older chips were rated in megahertz (MHz) and new chips in gigahertz (GHz). Since the dawn of the personal computer industry, motherboards have included oscillators, quartz crystals shaved down to a specific geometry so that engineers know exactly how they will react when a current is run through them. The phenomenon of a quartz crystal vibrating when exposed to a current is known as the piezoelectric effect. The crystal (XTL) known as the system clock keeps the time for the flow of data on the motherboard. How the front-side bus uses the clock leads to an effective clock rate known as the FSB speed. As discussed in the section “Types of Memory” later in this chapter, the FSB speed is computed differently for different types of RAM (DDR3, DDR4, DDR5, and so forth). From here, the CPU multiplies the FSB speed to produce its own internal clock rate, producing the third speed mentioned thus far.

As a result of the foregoing tricks of physics and mathematics, there can be a discrepancy between the front-side bus frequency and the internal frequency that the CPU uses to latch data and instructions through its pipelines. This disagreement between the numbers comes from the fact that the CPU is capable of splitting the clock signal it receives from the external oscillator that drives the front-side bus into multiple regular signals for its own internal use. In fact, you might be able to purchase a number of processors rated for different (internal) speeds that are all compatible with a single motherboard that has a front-side bus rated, for instance, at 1,333 MHz. Furthermore, you might be able to adjust the internal clock rate of the CPU that you purchased through settings in the BIOS.

The speed of a processor can also be tweaked by overclocking, or running the processor at a higher speed than the one at which the manufacturer rated it. Running at a higher speed requires more voltage and also generates more heat, which can shorten the life of the CPU. Manufacturers often discourage the practice (of course, they want you to just buy a faster and more expensive CPU), and it usually voids any warranty. However, some chips are sold today that specifically give you the ability to overclock them. Our official recommendation is to not do it unless the manufacturer says it's okay. If you're curious, plenty of information on how to overclock is available online.

Multithreading and Hyper-Threading

The string of instructions that a CPU runs is known as a thread. Old processors were capable of running only one thread at a time, whereas newer ones can run multiple threads at once. This is called multithreading.

Intel markets their multithreading technology as Hyper-Threading Technology (HTT). HTT is a form of simultaneous multithreading (SMT). SMT takes advantage of a modern CPU's superscalar architecture. Superscalar processors can have multiple instructions operating on separate data in parallel.

HTT-capable processors appear to the operating system to be two processors. As a result, the operating system can schedule two processes at the same time, as in the case of symmetric multiprocessing (SMP), where two or more processors use the same system resources. In fact, the operating system must support SMP in order to take advantage of HTT. If the current process stalls because of missing data caused by, say, cache or branch prediction issues, the execution resources of the processor can be reallocated for a different process that is ready to go, reducing processor downtime.

HTT manifests itself in the Windows 10 Task Manager by, for example, showing graphs for twice as many CPUs as the system has cores. These virtual CPUs are listed as logical processors (see Figure 1.28).

For an in-market example, compare the Intel i5 with the Intel i7. Similar models will have the same number of cores (say, four), but the i7 supports HTT, whereas the i5 does not. This gives the i7 a performance edge over its cousin. The i9 will be even one further step up from the i7. For everyday email and Internet use, the differences won't amount to much. But for someone who is using resource-intensive apps such as online gaming or virtual reality, the differences, especially between i5 and i9 processors, can be important.

Virtualization Support

Many of today's CPUs support virtualization in hardware, which eases the burden on the system that software-based virtualization imposes. For more information on virtualization, see Chapter 8, “Virtualization and Cloud Computing.” AMD calls its virtualization technology AMD-V (V for virtualization), whereas Intel calls theirs Virtualization Technology (VT). Most processors made today support virtual technology, but not all. Keep in mind that the BIOS/UEFI and operating system must support it as well for virtualization to work. You may need to manually enable the virtualization support in the BIOS/UEFI before it can be used. If you have an Intel processor and would like to check its support of VT, visit the following site to download the Intel Processor Identification Utility:

https://downloadcenter.intel.com/download/7838

As shown in Figure 1.30, the CPU Technologies tab of this utility tells you if your CPU supports Intel VT.

Parity Checking and Memory Banks

Parity checking is a rudimentary error-checking scheme that offers no error correction. Parity checking works most often on a byte, or 8 bits, of data. A ninth bit is added at the transmitting end and removed at the receiving end so that it does not affect the actual data transmitted. If the receiving end does not agree with the parity that is set in a particular byte, a parity error results. The four most common parity schemes affecting this extra bit are known as even, odd, mark, and space. Even and odd parity are used in systems that actually compute parity. Mark (a term for a digital pulse, or 1 bit) and space (a term for the lack of a pulse, or a 0 bit) parity are used in systems that do not compute parity but expect to see a fixed bit value stored in the parity location. Systems that do not support or reserve the location required for the parity bit are said to implement non-parity memory.

The most basic model for implementing memory in a computer system uses eight memory chips to form a set. Each memory chip holds millions or billions of bits of information, each in its own cell. For every byte in memory, one bit is stored in each of the eight chips. A ninth chip is added to the set to support the parity bit in systems that require it. One or more of these sets, implemented as individual chips or as chips mounted on a memory module, form a memory bank.

A bank of memory is required for the computer system to recognize electrically that the minimum number of memory components or the proper number of additional memory components has been installed. The width of the system data bus, the external bus of the processor, dictates how many memory chips or modules are required to satisfy a bank. For example, one 32-bit, 72-pin SIMM (single in-line memory module) satisfies a bank for an old 32-bit CPU, such as a i386 or i486 processor. Two such modules are required to satisfy a bank for a 64-bit processor—a Pentium, for instance. However, only a single 64-bit, 168-pin DIMM is required to satisfy the same Pentium processor. For those modules that have fewer than eight or nine chips mounted on them, more than 1 bit for every byte is being handled by some of the chips. For example, if you see three chips mounted, the two larger chips customarily handle 4 bits, a nibble, from each byte stored, and the third, smaller chip handles the single parity bit for each byte.

Even and odd parity schemes operate on each byte in the set of memory chips. In each case, the number of bits set to a value of 1 is counted up. If there is an even number of 1 bits in the byte (0, 2, 4, 6, or 8), even parity stores a 0 in the ninth bit, the parity bit; otherwise, it stores a 1 to even up the count. Odd parity does just the opposite, storing a 1 in the parity bit to make an even number of 1s odd and a 0 to keep an odd number of 1s odd. You can see that this is effective only for determining if there was a blatant error in the set of bits received, but there is no indication as to where the error is and how to fix it. Furthermore, the total 1-bit count is not important, only whether it's even or odd. Therefore, in either the even or odd scheme, if an even number of bits is altered in the same byte during transmission, the error goes undetected because flipping 2, 4, 6, or all 8 bits results in an even number of 1s remaining even and an odd number of 1s remaining odd.

Mark and space parity are used in systems that want to see 9 bits for every byte transmitted but don't compute the parity bit's value based on the bits in the byte. Mark parity always uses a 1 in the parity bit, and space parity always uses a 0. These schemes offer less error detection capability than the even and odd schemes because only changes in the parity bit can be detected. Again, parity checking is not error correction; it's error detection only, and not the best form of error detection at that. Nevertheless, an error can lock up the entire system and display a memory parity error. Enough of these errors and you need to replace the memory. Therefore, parity checking remains from the early days of computing as an effective indicator of large-scale memory and data-transmission failure, such as with serial interfaces attached to analog modems or networking console interfaces, but not so much for detecting random errors.

In the early days of personal computing, almost all memory was parity-based. As quality has increased over the years, parity checking in the RAM subsystem has become more uncommon. As noted earlier, if parity checking is not supported, there will generally be fewer chips per module, usually one less per column of RAM.

Error Checking and Correction

The next step in the evolution of memory error detection is known as error-correction code (ECC). If memory supports ECC, check bits are generated and stored with the data. An algorithm is performed on the data and its check bits whenever the memory is accessed. If the result of the algorithm is all zeros, then the data is deemed valid and processing continues. ECC can detect single- and double-bit errors and actually correct single-bit errors. In other words, if a particular byte—group of 8 bits—contains errors in 2 of the 8 bits, ECC can recognize the error. If only 1 of the 8 bits is in error, ECC can correct the error.

Single- and Double-Sided Memory

Commonly speaking, the terms single-sided memory and double-sided memory refer to how some memory modules have chips on one side and others have chips on both sides. Double-sided memory is essentially treated by the system as two separate memory modules. Motherboards that support such memory have memory controllers that must switch between the two “sides” of the modules and, at any particular moment, can access only the side to which they have switched. Double-sided memory allows more memory to be inserted into a computer, using half the physical space of single-sided memory, which requires no switching by the memory controller.

Single-, Dual-, Triple-, and Quad-Channel Memory

Standard memory controllers manage access to memory in chunks of the same size as the system bus's data width. This is considered communicating over a single channel. Most modern processors have a 64-bit system data bus. This means that a standard memory controller can transfer exactly 64 bits of information at a time. Communicating over a single channel is a bottleneck in an environment where the CPU and memory can both operate faster than the conduit between them. Up to a point, every channel added in parallel between the CPU and RAM serves to ease this constriction.

Memory controllers that support dual-channel and greater memory implementation were developed in an effort to alleviate the bottleneck between the CPU and RAM. Dual-channel memory is the memory controller's coordination of two memory banks to work as a synchronized set during communication with the CPU, doubling the specified system bus width from the memory's perspective. Triple-channel memory, then, demands the coordination of three memory modules at a time. Quad-channel memory is the coordination of four memory modules at once. Collectively, they are known as multichannel memory implementations.

Because today's processors largely have 64-bit external data buses, and because one stick of memory satisfies this bus width, there is a 1:1 ratio between banks and modules. This means that implementing multichannel memory in today's most popular computer systems requires that two, three, or four memory modules be installed at a time. Note, however, that it's the motherboard, not the memory, that implements multichannel memory (more on this in a moment). Single-channel memory, in contrast, is the classic memory model that dictates only that a complete bank be satisfied whenever memory is initially installed or added. One bank supplies only half the width of the effective bus created by dual-channel support, for instance, which by definition pairs two banks at a time.

In almost all cases, multichannel implementations support single-channel installation, but poorer performance should be expected. Multichannel motherboards often include slots of different colors, usually one of each color per set of slots. To use only a single channel, you populate slots of the same color, skipping neighboring slots to do so. Filling neighboring slots in a dual-channel motherboard takes advantage of its dual-channel capability.

Because of the special tricks that are played with memory subsystems to improve overall system performance, care must be taken during the installation of disparate memory modules. In the worst case, the computer will cease to function when modules of different speeds, different capacities, or different numbers of sides are placed together in slots of the same channel. If all of these parameters are identical, there should be no problem with pairing modules. Nevertheless, problems could still occur when modules from two different manufacturers or certain unsupported manufacturers are installed, all other parameters being the same. Technical support or documentation from the manufacturer of your motherboard should be able to help with such issues.

Although it's not the make-up of the memory that leads to multichannel support but instead the technology on which the motherboard is based, some memory manufacturers still package and sell pairs and triplets of memory modules in an effort to give you peace of mind when you're buying memory for a system that implements multichannel memory architecture. Keep in mind, the motherboard memory slots have the distinctive color-coding, not the memory modules.

DRAM

DRAM is dynamic random access memory. This is what most people are talking about when they mention RAM. When you expand the memory in a computer, you are adding DRAM chips. You use DRAM to expand the memory in the computer because it's a cheaper type of memory. Dynamic RAM chips are cheaper to manufacture than most other types because they are less complex. Dynamic refers to the memory chips' need for a constant update signal (also called a refresh signal) in order to keep the information that is written there. If this signal is not received every so often, the information will bleed off and cease to exist. Currently, the most popular implementations of DRAM are based on synchronous DRAM and include DDR3 and DDR4. Occasionally you will see some DDR2, and DDR5 is new so it hasn't been widely adopted yet. Before discussing these technologies, let's take a quick look at the legacy asynchronous memory types, none of which should appear on modern exams.

SRAM

Static random access memory (SRAM) doesn't require a refresh signal like DRAM does. The chips are more complex and are thus more expensive. However, they are considerably faster. DRAM access times come in at 40 nanoseconds (ns) or more; SRAM has access times faster than 10ns. SRAM is classically used for cache memory.

ROM

ROM stands for read-only memory. It is called read-only because you could not write to the original form of this memory. Once information had been etched on a silicon chip and manufactured into the ROM package, the information couldn't be changed. Some form of ROM is normally used to store the computer's BIOS because this information normally does not change often.

The system ROM in the original IBM PC contained the power-on self-test (POST), BIOS, and cassette BASIC. Later, IBM computers and compatibles included everything but the cassette BASIC. The system ROM enables the computer to “pull itself up by its bootstraps,” or boot (find and start the operating system).

Through the years, different forms of ROM were developed that could be altered, later ones more easily than earlier ones. The first generation was the programmable ROM (PROM), which could be written to for the first time in the field using a special programming device, but then no more. You may liken this to the burning of a DVD-R.

The erasable PROM (EPROM) followed the PROM, and it could be erased using ultraviolet light and subsequently reprogrammed using the original programming device. These days, flash memory is a form of electronically erasable PROM (EEPROM). Of course, it does not require UV light to erase its contents, but rather a slightly higher than normal electrical pulse.

DIMM

One type of memory package is known as a DIMM, which stands for dual in-line memory module. DIMMs are 64-bit memory modules that are used as a package for the SDRAM family: SDR, DDR, DDR2, DDR3, DDR4, and DDR5. The term dual refers to the fact that, unlike their SIMM predecessors, DIMMs differentiate the functionality of the pins on one side of the module from the corresponding pins on the other side. With 84 pins per side, this makes 168 independent pins on each standard SDR module, as shown with its two keying notches as well as the last pin labeled 84 on the right side in Figure 1.33. SDR SDRAM modules are no longer part of the CompTIA A+ objectives, and they are mentioned here as a foundation only.

The DIMM used for DDR memory has a total of 184 pins and a single keying notch, whereas the DIMM used for DDR2 has a total of 240 pins, one keying notch, and possibly an aluminum cover for both sides, called a heat spreader and designed like a heat sink to dissipate heat away from the memory chips and prevent overheating. The DDR3 DIMM is similar to that of DDR2. It has 240 pins and a single keying notch, but the notch is in a different location to avoid cross-insertion. Not only is the DDR3 DIMM physically incompatible with DDR2 DIMM slots, it's also electrically incompatible. A DDR4 DIMM is the same length as a DDR3 DIMM, but is about 0.9mm taller and has 288 pins. The key is in a different spot, so you can't put DDR4 memory into a DDR2 or DDR3 slot. Finally, DDR5 has 288 pins as DDR4 does but is keyed differently so that DDR4 modules won't fit into DDR5 slots, and vice versa. Table 1.3 summarizes some key differences between the types of DDR we've introduced in this chapter.

Figure 1.34 shows, from top to bottom, DDR4, DDR3, and DDR2 DIMMs.

SODIMM

Laptop computers and other computers that require much smaller components don't use standard RAM packages, such as DIMMs. Instead, they call for a much smaller memory form factor, such as a small outline DIMM (SODIMM). SODIMMs are available in many physical implementations, including the older 32-bit (72- and 100-pin) configuration and newer 64-bit (144-pin SDR SDRAM, 200-pin DDR/DDR2, 204-pin DDR3, 260-pin DDR4, and 262-pin DDR5) configurations.

All 64-bit modules have a single keying notch. The 144-pin module's notch is slightly off center. Note that although the 200-pin SODIMMs for DDR and DDR2 have slightly different keying, it's not so different that you don't need to pay close attention to differentiate the two. They are not, however, interchangeable. DDR3, DDR4, and DDR5 are keyed differently from the others as well. Figure 1.34 shows a DDR3 SODIMM compared to DDR3 and DDR2 DIMMs.

Air Cooling

The parts inside most computers are cooled by air moving through the case. The CPU is no exception. However, because of the large amount of heat produced, the CPU must have (proportionately) the largest surface area exposed to the moving air in the case. Therefore, the heat sinks on the CPU are the largest of any inside the computer.

The CPU fan often blows air down through the body of the heat sink to force the heat into the ambient internal air where it can join the airflow circuit for removal from the case. However, in some cases, you might find that the heat sink extends up farther, using radiator-type fins, and the fan is placed at a right angle and to the side of the heat sink. This design moves the heat away from the heat sink immediately instead of pushing the air down through the heat sink. CPU fans can be purchased that have an adjustable rheostat to allow you to dial in as little airflow as you need, aiding in noise reduction but potentially leading to accidental overheating.

It should be noted that the highest-performing CPU coolers use copper plates in direct contact with the CPU. They also use high-speed and high-CFM cooling fans to dissipate the heat produced by the processor. CFM is short for cubic feet per minute, an airflow measurement of the volume of air that passes by a stationary object per minute. Figure 1.41 shows a newer, large heat sink with a fan in the center. In the picture it can be tough to gauge size—this unit is about six inches across! (And to be fair, this heat sink should have a second fan on one of the sides, but with the second fan the heatsink wouldn't fit into the author's case—the RAM was in the way.)

Most new CPU heat sinks use tubing to transfer heat away from the CPU. With any cooling system, the more surface area exposed to the cooling method, the better the cooling. Plus, the heat pipes can be used to transfer heat to a location away from the heat source before cooling. This is especially useful in cases where the form factor is small and with laptops, where open space is limited.

With advanced heat sinks and CPU cooling methods like this, it is important to improve the thermal transfer efficiency as much as possible. To that end, cooling engineers came up with a glue-like compound that helps to bridge the extremely small gaps between the CPU and the heat sink, which avoids superheated pockets of air that can lead to focal damage of the CPU. This product is known as thermal transfer compound, or simply thermal compound (alternatively, thermal grease or thermal paste), and it can be bought in small tubes. Single-use tubes are also available and alleviate the guesswork involved with how much you should apply. Watch out, though; this stuff makes quite a mess and doesn't want to come off your fingers very easily. An alternative to the paste is a small thermal pad, which provides heat conductivity between the processor and the heat sink.

Apply the compound by placing a bead in the center of the heat sink, not on the CPU, because some heat sinks don't cover the entire CPU package. That might sound like a problem, but some CPUs don't have heat-producing components all the way out to the edges. Some CPUs even have a raised area directly over the silicon die within the packaging, resulting in a smaller contact area between the components. You should apply less than you think you need because the pressure of attaching the heat sink to the CPU will spread the compound across the entire surface in a very thin layer. It's advisable to use a clean, lint-free applicator of your choosing to spread the compound around a bit as well, just to get the spreading started. You don't need to concern yourself with spreading it too thoroughly or too neatly because the pressure applied during attachment will equalize the compound quite well. During attachment, watch for oozing compound around the edges, clean it off immediately, and use less next time.

If you've ever installed a brand-new heat sink onto a CPU, you've most likely used thermal compound or the thermal compound patch that was already applied to the heat sink for you. If your new heat sink has a patch of thermal compound pre-applied, don't add more. If you ever remove the heat sink, don't try to reuse the patch or any other form of thermal compound. Clean it all off and start fresh.

Liquid Cooling

Liquid cooling is a technology whereby a special water block is used to conduct heat away from the processor (as well as from the chipset). Water is circulated through this block to a radiator, where it is cooled.

The theory is that you could achieve better cooling performance through the use of liquid cooling. For the most part, this is true. However, with traditional cooling methods (which use air and water), the lowest temperature you can achieve is room temperature. Plus, with liquid cooling, the pump is submerged in the coolant (generally speaking), so as it works, it produces heat, which adds to the overall liquid temperature.

The main benefit to liquid cooling is silence. Only one fan is needed: the fan on the radiator to cool the water. So, a liquid-cooled system can run extremely quietly.

There are two major classifications of liquid cooling systems in use with PCs today: all-in-one (AIO) coolers and custom loop systems. AIO systems are relatively easy to install—they require about as much effort as a heat sink and fan—and comparably priced to similarly effective air systems. Figure 1.42 shows an example from Corsair, with the pump in front and the fans behind it, attached to the radiator.

AIO systems come in three common sizes: 120 mm (with one fan, and the most common), 240 mm (two fans, for overclocked components), and 360 mm (three fans, for high-end multicore overclocked components). Options with RGB lighting are readily available if that's the style you want. Custom loop systems can quickly become complex and expensive, but many hardcore gamers swear by their performance. The components are essentially the same as those in an AIO system—there's a radiator, pump, fans, some tubes, and liquid. However, each part is purchased separately, and some assembly is required.

Sound Cards

Just as there are devices to convert computer signals into printouts and video information, there are devices to convert those signals into sound. These devices are known as sound cards. Although sound cards started out as pluggable adapters, this functionality is one of the most common integrated technologies found on motherboards today. A sound card typically has small, round 1/8 jacks on the back of it for connecting microphones, headphones, and speakers as well as other sound equipment. Older sound cards used a DA15 game port, which could be used for either joysticks or Musical Instrument Digital Interface (MIDI) controllers. Figure 2.2 shows an example of a sound card with a DA15 game port.

In our section on video cards, we noted that integrated cards have inferior performance to add-on ones, and though the same holds true for sound cards, the difference isn't quite as drastic. Many of today's motherboards come equipped with 5.1 or 7.1 analog or digital audio and support other surround sound formats as well. For everyday users and even many gamers, integrated audio is fine.

For users who need extra juice, such as those who produce movies or videos or do other audio/video (A/V) editing, a specialized add-on sound card is a must. Very good quality sound cards can be found for under $100, compared with cheaper models around $20—there isn't a huge difference in price as there is with CPUs and video cards. Look for a card with a higher sampling rate (measured in kilohertz [kHz]) and higher signal-to-noise ratio (measured in decibels [dB]). The de facto standard for sound cards is the Sound Blaster brand. Although other brands exist, they will often tout “Sound Blaster compatibility” in their advertising to show that they are legit.

In addition to audio output, many A/V editors will require the ability to input custom music from an electronic musical keyboard or other device. A term you will hear in relation to this is the MIDI standard. As noted earlier, old sound cards would sometimes have a round 5-pin MIDI port, which was used to connect the musical keyboard or other instrument to the computer. Today, digital musical instrument connections are often made via USB. Nonetheless, you will still see the term MIDI compatible used with a lot of digital musical devices.

Video Capture Cards

A video capture card is a stand-alone add-on card often used to save a video stream to the computer for later manipulation or sharing. This can be video from an Internet site, or video from an external device such as a digital camera or smartphone. Video-sharing sites on the Internet make video capture cards quite popular with enterprises and Internet users alike. Video capture cards need and often come with software to aid in the processing of multimedia input. While video and sound cards are internal expansion devices, capture cards can be internal (PCIe) or external (USB).

Not all video capture cards record audio signals while processing video signals. If this feature is important to you or the end user, be sure to confirm that the card supports it. Also know that capture cards work with standard video resolutions, and specific cards might be limited on the resolutions they support. Double-check the specifications to make sure the card will meet the need, and also make sure to get reviews of the software used with the device.

Anatomy of a Hard Drive

A hard drive is constructed in a cleanroom to avoid the introduction of contaminants into the hermetically sealed drive casing. Once the casing is sealed, most manufacturers seal one or more of the screws with a sticker warning that removal of or damage to the seal will result in voiding the drive's warranty. Even some of the smallest contaminants can damage the precision components if allowed inside the hard drive's external shell. The following is a list of the terms used to describe these components in the following paragraphs:

• Platters
• Read/write heads
• Tracks
• Sectors
• Cylinders
• Clusters (allocation units)

Inside the sealed case of the hard drive lie one or more platters, where the actual data is stored by the read/write heads. The heads are mounted on a mechanism that moves them in tandem across both surfaces of all platters. Older drives used a stepper motor to position the heads at discrete points along the surface of the platters, which spin at thousands of revolutions per minute on a spindle mounted to a hub. Newer drives use voice coils for a more analog movement, resulting in reduced data loss because the circuitry can sense where the data is located through a servo scheme, even if the data shifts due to changes in physical disc geometry. Figure 2.10 shows the internal components of a conventional hard drive.

Before a hard drive can store data, it must be prepared. Factory preparation for newer drives, or low-level formatting in the field for legacy drives, maps the inherent flaws of the platters so that the drive controllers know not to place data in these compromised locations. Additionally, this phase in drive preparation creates concentric rings, or tracks, which are drawn magnetically around the surface of the platters. Sectors are then delineated within each of the tracks. Sectors are the magnetic domains that represent the smallest units of storage on the disk's platters. This is illustrated in Figure 2.11. Magnetic-drive sectors commonly store only 512 bytes (½ KB) of data each.

The capacity of a hard drive is a function of the number of sectors it contains. The controller for the hard drive knows exactly how the sectors are laid out within the disk assembly. It takes direction from the BIOS when writing information to and reading information from the drive. The BIOS, however, does not always understand the actual geometry of the drive. For example, the BIOS does not support more than 63 sectors per track. Nevertheless, almost all hard drives today have tracks that contain many more than 63 sectors per track. As a result, a translation must occur from where the BIOS believes it is directing information to be written to where the information is actually written by the controller. When the BIOS detects the geometry of the drive, it is because the controller reports dimensions that the BIOS can understand. The same sort of trickery occurs when the BIOS reports to the operating system a linear address space for the operating system to use when requesting that data be written to or read from the drive through the BIOS.

After initial drive preparation, the drive is formatted with a file system, by the operating system, and then it's ready to store data. Filesystems laid down on the tracks and their sectors routinely group a configurable number of sectors into equal or larger sets called clusters or allocation units. This concept exists because operating system designers have to settle on a finite number of addressable units of storage and a fixed number of bits to address them uniquely.

No two files are allowed to occupy the same sector, so the opportunity exists for a waste of space if small files occupy only part of a sector. Clusters exacerbate the problem by having a similar foible: the operating system does not allow any two files to occupy the same cluster. Thus, the larger the cluster size, the larger the potential waste. So although you can increase the cluster size (generally to as large as 64 KB, which corresponds to 128 sectors), you should keep in mind that unless you are storing a notable number of very large files, the waste will escalate astoundingly, perhaps negating or reversing your perceived storage-capacity increase.

HDD Speeds

As the electronics within the HBA and controller get faster, they are capable of requesting data at higher and higher rates. If the platters are spinning at a constant rate, however, the information can be accessed only as fast as a given fixed rate. To make information available to the electronics more quickly, manufacturers increase the speed at which the platters spin from one generation of drives to the next, with multiple speeds coexisting in the marketplace for an unpredictable period, at least until the demand dies down for one or more speeds.

The following spin rates have been used in the industry for the platters in conventional magnetic hard disk drives:

• 5,400 rpm
• 7,200 rpm
• 10,000 rpm
• 12,000 rpm
• 15,000 rpm

While it is true that a higher revolutions per minute (rpm) rating results in the ability to move data more quickly, there are many applications that do not benefit from increased disk-access speeds. As a result, you should choose only faster drives, which are also usually more expensive per byte of capacity, when you have an application for this type of performance, such as for housing the partition where the operating system resides or for very disk-intensive programs. For comparison, a 7,200 rpm SATA hard drive can sustain data read speeds of about 100 MBps, which is about the same as a PATA ATA/100 7,200 rpm drive. A 10,000 rpm (also known as 10k) SATA drive can top out around 200 MBps.

Higher speeds also consume more energy and produce more heat. The lower speeds can be ideal in laptops, where heat production and battery usage can be issues with higher-speed drives. Even the fastest conventional hard drives are slower than solid-state drives are at transferring data.

HDD Form Factors

Physically, the most common hard drive form factors (sizes) are 3.5" and 2.5". Desktops traditionally use 3.5" drives, whereas the 2.5" drives are made for laptops—although most laptops today avoid using conventional HDDs. Converter kits are available to mount a 2.5" drive into a 3.5" desktop hard drive bay. Figure 2.12 shows the two drives together. As you can see, the 2.5" drive is significantly smaller in all three dimensions, but it does have the same connectors as its bigger cousin.

Hybrid Drives

A cost-saving alternative to a standard SSD that can still provide a significant increase in performance over conventional HDDs is the hybrid drive. Hybrid drives can be implemented in two ways: a solid-state hybrid drive and a dual-drive storage solution. Both forms of hybrid drives can take advantage of solutions such as Intel's Smart Response Technology (SRT), which informs the drive system of the most used and highest-value data. The drive can then load a copy of such data into the SSD portion of the hybrid drive for faster read access.

It should be noted that systems on which data is accessed randomly do not benefit from hybrid drive technology. Any data that is accessed for the first time will also not be accessed from flash memory, and it will take as long to access it as if it were accessed from a traditional hard drive. Repeated use, however, will result in the monitoring software's flagging of the data for caching in the SSD.

SSD Communication Interfaces

It's been said that the advent of the SSD was a major advancement for the computer industry. Solid-state drives are basically made from the same circuitry that RAM is, and they are really, really fast. When they were first on the market, the limitation in the system was the SATA controller that most hard drives were plugged into. So as enterprising computer engineers are known to do, some started looking into ways to overcome this barrier. The result is that there have been more interfaces designed for storage devices, with much faster speeds.

The CompTIA A+ exam objectives list three technologies as SSD communications interfaces: SATA, PCIe, and NVMe. We will cover them now.

PCIe

Peripheral Component Interconnect Express (PCIe) is another technology that was covered in Chapter 1 and is used for SSDs. PCIe was first introduced in 2002, technically a year before SATA, and both technologies took a little while to get widely adopted. Like SATA, PCIe has gone through several revisions, with each version being faster than the previous one. Table 2.1 shows the throughput of PCIe versions.

Version	Transfer rate	Throughput per lane (one direction)	Total x16 throughput (bidirectional)
1.0	2.5 GTps	250 MBps	8 GBps
2.0	5.0 GTps	500 MBps	16 GBps
3.0	8.0 GTps	1 GBps	32 GBps
4.0	16.0 GTps	2 GBps	64 GBps
5.0	32.0 GTps	4 GBps	128 GBps

Looking at Table 2.1, something might immediately jump out at you. The transfer rate is specified in gigatransfers per second (GTps). This unit of measure isn't very commonly used, and most people talk about PCIe in terms of its data throughput. Before moving on, though, let's take a quick look at what GTps means. (It's highly unlikely you will be tested on this, but you might be curious.)

PCIe is a serial bus that embeds clock data into the data stream to help the sender and receiver keep track of the order of transmissions. Two clock bits are used for every eight bits of data. Therefore, PCIe uses what's called “8b/10b” encoding—8 bits of data are sent in a 10-bit bundle, which is decoded at the receiving end. Gigatransfers per second refers to the total number of bits sent, whereas the throughput data you might be more used to seeing refers to data only and not the clock.

Also remember that PCIe cards can have different numbers of channels—x1, x2, x4, x8, and x16. So, for example, a PCIe 3.0 x4 card will have total data throughput of 8 GBps (1 GBps for one lane in one direction, times two for bidirectional, times four for the four channels).

What does all of this mean for SSDs? First, take a look at a picture of a PCIe SSD in Figure 2.13. This is a PCIe 2.0 x4 Kingston HyperX Predator. These drives came in capacities up to 960 GB and supported data reads of up to 2.64 GBps and maximum write speeds of 1.56 GBps. The drive is a little dated by today's standards, but it still serves as a great example. First, it's the most common PCIe SSD size, which is x4. PCIe x2 drives are also common, with x8 and x16 drives being relatively rare. Second, notice the transfer speeds. Based on Table 2.1, PCIe 2.0 x4 has a maximum throughput of 4 GBps. This drive doesn't get to that level, especially on write speeds. This is the difference between theoretical maximums of standards versus practical realities of creating hardware. Even so, you can see that this PCIe SSD is significantly faster than a SATA SSD.

NVMe

Created by a consortium of manufacturers, including Intel, Samsung, Dell, SanDisk, and Seagate, and released in 2011, Non-Volatile Memory Express (NVMe) is an open standard designed to optimize the speed of data transfers. Unlike SATA and PCIe, NVMe isn't related to a specific type of connector. Said another way, there is no NVMe connector—think of it as a nonvolatile memory chip that can be used in SATA, PCIe, or M.2 (which we will cover in the next section) slots. Figure 2.14 shows a 1 TB Western Digital NVMe SSD.

NVMe drives are frighteningly fast—current NVMe SSDs can support data reads of up to 3.5 GBps, provided that the interface they are plugged into supports it as well, of course. An NVMe SATA 3 SSD will still be limited to 600 MBps; older PCIe versions in theory might have limitations, but you're not going to find any PCIe 1.0 or 2.0 NVMe SSDs, so it's not a problem.

One potential issue you might see with NVMe hard drives is that in order to be a boot drive, the motherboard must support it. If the motherboard has a built-in M.2 slot, odds are that the BIOS will support booting from an NVMe drive. If you are adding it using a PCIe port, the BIOS might not be able to boot from that drive. Always check the motherboard documentation to ensure it supports what you're trying to do.

SSD Form Factors

Whereas a communications interface is the method the device uses to communicate with other components, a form factor describes the shape and size of a device. The two SSD form factors you need to know for the A+ exam are mSATA and M.2.

mSATA

The Serial ATA International Organization has developed several specifications—you've already been introduced to SATA and eSATA. Next on the list is a form factor specifically designed for portable devices such as laptops and smaller—mini-Serial ATA (mSATA). mSATA was announced in 2009 as part of SATA version 3.1 and hit the market in 2010.

mSATA uses the same physical layout as the Mini PCI Express (mPCIe) standard, and both have a 30 mm-wide 52-pin connector. The wiring and communications interfaces between the two standards are different though. mSATA uses SATA technology, whereas mPCIe uses PCIe. In addition, mPCIe card types are as varied as their larger PCIe cousins are, including video, network, cellular, and other devices. mSATA, on the other hand, is dedicated to storage devices based on SATA bus standards. mSATA cards come in 30 mm × 50.95 mm full-size and 30mm × 26.8 mm half-size cards. Figure 2.15 shows a full-sized mSATA SSD on top of a 2.5" SATA SSD.

The wiring differences between mSATA and mPCIe can pose some interesting challenges. Both types of cards will fit into the same slot, but it depends on the motherboard as to which is supported. You might have heard us say this before, but as always, check the motherboard's documentation to be sure.

M.2

Originally developed under the name Next Generation Form Factor (NGFF), M.2 (pronounced “M dot 2”) was born from the desire to standardize small form factor SSD hard drives. We touched briefly on M.2 in Chapter 1, and we mentioned that although M.2 is primarily used for hard drives, it supports other types of cards such as Wi-Fi, Bluetooth, Global Positioning System (GPS), and near-field communication (NFC) connectivity, as well as PCIe and SATA connections. It's a form factor designed to replace the mSATA standard for ultra-small expansion components in laptops and smaller devices. Whereas mSATA uses a 30mm 52-pin connector, M.2 uses a narrower 22 mm 66-pin connector.

One interesting connectivity feature of M.2 is that the slots and cards are keyed such that only a specific type of card can fit into a certain slot. The keys are given letter names to distinguish them from each other, starting with the letter A and moving up the alphabet as the location of the key moves across the expansion card. Table 2.2 explains the slot names, some interface types supported, and common uses.

Key	Common interfaces	Uses
A	PCIe x2, USB 2.0	Wi-Fi, Bluetooth, and cellular cards
B	PCIe x2, SATA, USB 2.0, USB 3.0, audio	SATA and PCIe x2 SSDs
E	PCIe x2, USB 2.0	Wi-Fi, Bluetooth, and cellular cards
M	PCIe x4, SATA	PCIe x4 SSDs

Let's look at some examples. Figure 2.16 shows four different M.2 cards. From left to right, they are an A- and E-keyed Wi-Fi card, two B- and M-keyed SSDs, and an M-keyed SSD. Of the four, only the M-keyed SSD can get the fastest speeds (up to 1.8 GBps), because it supports PCIe x4. SSDs on the market are keyed B, M, or B+M. A B-keyed or M-keyed SSD won't fit in a B+M socket. A B+M keyed drive will fit into a B socket or an M socket, however.

Another interesting feature of the cards is that they are also named based on their size. For example, you will see card designations such as 1630, 2242, 2280, 22110, or 3042. The first two numbers refer to the width, and the rest to the length (in millimeters) of the card. In Figure 2.16, you see a 1630, a 2242, and two 2280 cards.

Figure 2.17 shows a motherboard with two M.2 slots. The one on the left is E-keyed, and the one on the right is B-keyed. The left slot is designed for an E-keyed Wi-Fi NIC, and the right one for a B-keyed SSD.

Many motherboards today come with protective covers over the M.2 slots. Adding these covers to provide a bit of safety within the case is a welcome feature. An example is shown in Figure 2.18. The bottom M.2 slot is covered, and the top slot (just above the PCIe x4 connector) has the cover removed. Notice the screw holes to support 42 mm, 60 mm, 80 mm, and 110 mm length devices.

As mentioned earlier, M.2 is a form factor, not a bus standard. M.2 supports SATA, USB, and PCIe buses. What does that mean for M.2 hard drives? It means that if you purchase an M.2 SATA hard drive, it will have the same speed limitation as SATA III, or about 600 MBps. That's not terrible, but it means that the primary advantage of an M.2 SATA drive versus a conventional SATA SSD is size. An M.2 PCIe hard drive is an entirely different story. PCIe, you will recall, is much faster than SATA. A PCIe 2.0 x1 bus supports one-way data transfers of 500 MBps. That is close to SATA III speed, and it's only a single lane for an older standard. NVMe M.2 drives kick up the speed even further. If you want the gold standard for hard drive speed, NVMe M.2 is the way to go.

To wrap up this section on SSDs, let's look at one more picture showing the difference in sizes between a few options, all from the manufacturer Micron. Figure 2.19 has a 2.5" SSD on top, with (from left to right) a full-sized mSATA drive, an M.2 22110 drive, and an M.2 2280 drive. All are SSDs that can offer the same capacity, but the form factors differ quite a bit from each other.

Flash Memory

Once used only for primary memory, the same components that sit on your motherboard as RAM can be found in various physical sizes and quantities among today's solid-state storage solutions. These include older removable and nonremovable flash memory mechanisms, Secure Digital (SD) and other memory cards, and USB flash drives. Each of these technologies has the potential to store reliably a staggering amount of information in a minute form factor. Manufacturers are using innovative packaging for some of these products to provide convenient transport options (such as keychain attachments) to users. Additionally, recall the SSD alternatives to magnetic hard drives mentioned earlier in this chapter.

For many years, modules known as flash memory have offered low- to mid-capacity storage for devices. The name comes from the concept of easily being able to use electricity to alter the contents of the memory instantly. The original flash memory is still used in devices that require a nonvolatile means of storing critical data and code often used in booting the device, such as routers and switches.

For example, Cisco Systems uses flash memory in various forms to store its Internetwork Operating System (IOS), which is accessed from flash during bootup and, in certain cases, throughout operation uptime and therefore during an administrator's configuration sessions. Lesser models store the IOS in compressed form in the flash memory device and then decompress the IOS into RAM, where it is used during configuration and operation. In this case, the flash memory is not accessed again after the boot-up process is complete, unless its contents are being changed, as in an IOS upgrade. Certain devices use externally removable PC Card technology as flash memory for similar purposes.

The following sections explain a bit more about today's most popular forms of flash memory: USB flash drives and memory cards.

USB Flash Drives

USB flash drives are incredibly versatile and convenient devices that enable you to store large quantities of information in a very small form factor. Many such devices are merely extensions of the host's USB connector, extending out from the interface but adding little to its width, making them easy to transport, whether in a pocket or a laptop bag. Figure 2.21 illustrates an example of one of these components and its relative size.

USB flash drives capitalize on the versatility of the USB interface, taking advantage of Windows' Plug and Play, AutoPlay, and Safely Remove Hardware features and the physical connector strength. Upon insertion, these devices announce themselves to Windows File Explorer as removable drives, and they show up in the Explorer window with a drive letter. This software interface allows for drag-and-drop copying and most of the other Explorer functions performed on standard drives. Note that you might have to use the Disk Management utility (discussed in Chapter 13) to assign a drive letter manually to a USB flash drive if it fails to acquire one itself. This can happen in certain cases, such as when the previous letter assigned to the drive has been taken by another device in the USB flash drive's absence.

SD and Other Memory Cards

Today's smaller devices require some form of removable solid-state memory that can be used for temporary and permanent storage of digital information. Modern electronics, as well as most contemporary digital still cameras, use some form of removable memory card to store still images permanently or until they can be copied off or printed out. Of these, the Secure Digital (SD) format has emerged as the preeminent leader of the pack, which includes the older MultiMediaCard (MMC) format on which SD is based. Both of these cards measure 32 mm × 24 mm, and slots that receive them are often marked for both. The SD card is slightly thicker than the MMC and has a write-protect notch (and often a switch to open and close the notch), unlike MMC.

Even smaller devices, such as mobile phones, have an SD solution for them. One of these products, known as miniSD, is slightly thinner than SD and measures 21.5 mm × 20 mm. The other, microSD, is thinner yet and only 15 mm × 11 mm. Both of these reduced formats have adapters that allow them to be used in standard SD slots. Figure 2.22 shows an SD card and a microSD card next to a ruler based on inches.

Table 2.3 lists additional memory card formats, the slots for some of which can be seen in the images that follow the table.

Format	Dimensions	Details	Year introduced
CompactFlash (CF)	36 mm × 43 mm	Type I and Type II variants; Type II used by IBM for Microdrive	1994
xD-Picture Card	20 mm × 25 mm	Used primarily in digital cameras	2002

Figure 2.23 shows the memory-card slots of an HP PhotoSmart printer, which is capable of reading these devices and printing from them directly or creating a drive letter for access to the contents over its USB connection to the computer. Clockwise from the upper left, these slots accommodate CF/Microdrive, SmartMedia, Memory Stick (bottom right), and MMC/SD. The industry provides almost any adapter or converter to allow the various formats to work together.

Nearly all of today's laptops have built-in memory card slots and many desktops will have readers built into the front or top panel of the case as well. If a computer doesn't have memory card slots built into the case, it's easy to add external card readers. Most are connected via USB, such as the one shown in Figure 2.24 (front first, then back), and are widely available in many different configurations.

Hot-Swappable Devices

Many of the removable storage devices mentioned are hot-swappable. This means that you can insert and remove the device with the system powered on. Most USB-attached devices without a filesystem fall into this category. Non–hot-swappable devices, in contrast, either cannot have the system's power applied when they are inserted or removed or have some sort of additional conditions for their insertion or removal. One subset is occasionally referred to as cold-swappable, the other as warm-swappable. The system power must be off before you can insert or remove cold-swappable devices. An example of a cold-swappable device is anything connected to a SATA connector on the motherboard.

Warm-swappable devices include USB flash drives and external drives that have a filesystem. Windows and other operating systems tend to leave files open while accessing them and write cached changes to them at a later time, based on the algorithm in use by the software. Removing such a device without using the Safely Remove Hardware and Eject Media utility can result in data loss. However, after stopping the device with the utility, you can remove it without powering down the system—hence, the warm component of the category's name. These are officially hot-swappable devices.

Hardware-based RAID systems benefit from devices and bays with a single connector that contains both power and data connections instead of two separate connectors. This is known as Single Connector Attachment (SCA). SCA interfaces have ground leads that are longer than the power leads so that they make contact first and lose contact last. SATA power connectors are designed in a similar fashion for the same purpose. This arrangement ensures that no power leads make contact without their singular ground leads, which would often result in damage to the drive. Drives based on SCA are hot-swappable. RAID systems that have to be taken offline before drives are changed out, but the system power can remain on, are examples of warm-swappable systems.

Optical Drives

The final category of storage devices we will look at is optical drives. They get their name because instead of storing data using magnetic fields like conventional HDDs, they read and write data with the use of a laser. The laser scans the surface of a spinning plastic disc, with data encoded as small bits and bumps in the track of the disc.

With the popularity of high-speed Internet access and streaming services, optical drives have lost much of their popularity. For about 20 years they were practically required components, but today they are far less common. The most advanced optical storage technology used is the Blu-ray Disc (BD) drive. It replaced the digital versatile disc (DVD), also called digital video disc drive, which in turn replaced the compact disc (CD) drive. Each type of optical drive can also be expected to support the technology that came before it. Such optical storage devices began earnestly replacing floppy diskette drives in the late 1990s. Although, like HDDs, these discs have greater data capacity and increased performance over floppies, they are not intended to replace hard disk drives. HDDs greatly exceed the capacity and read/write performance of optical drives.

The CDs, DVDs, and BDs used for data storage, which may require multiple data reads and writes, are virtually the same as those used for permanent recorded audio and video. The way that data, audio, and video information is written to consumer-recordable versions makes them virtually indistinguishable from such professionally manufactured discs. Any differences that arise are due to the format used to encode the digital information on the disc. Despite the differences among the professional and consumer formats, newer players have no issue with any type of disc used. Figure 2.25 shows an example of an internal 5¼" DVD-ROM drive, which also accepts CD-ROM discs. Modern optical drives are indistinguishable from older ones, aside from obvious markings concerning their capabilities. External drives that connect via USB are more popular (and portable!) than their internal cousins.

Removing Storage Devices

Removing any component is frequently easier than installing the same part. Consider the fact that most people could destroy a house, perhaps not safely enough to ensure their well-being, but they don't have to know the intricacies of construction to start smashing away. On the other hand, very few people are capable of building a house. Similarly, many could figure out how to remove a storage device, as long as they can get into the case to begin with, but only a few could start from scratch and successfully install one without tutelage.

In Exercise 2.1, you'll remove an internal storage device.

Installing Storage Devices

An obvious difference among storage devices is their form factor. This is the term used to describe the physical dimensions of a storage device. Form factors commonly have the following characteristics:

• 3¼" wide vs. 5¼" wide
• Half height vs. full height vs. 1" high and more
• Any of the laptop specialty form factors

You will need to determine whether you have an open bay in the chassis to accommodate the form factor of the storage device that you want to install. Adapters exist that allow a device of small size to fit into a larger bay. For obvious reasons, the converse is not also true.

In Exercise 2.2, you'll install an internal storage device.

ATX, ATX12V, and EPS12V Connectors

ATX motherboards use a single block connector from the power supply. When ATX boards were first introduced, this connector was enough to power all the motherboard, CPU, memory, and all expansion slots. The original ATX system connector provides the six voltages required, plus it delivers them all through one connector: an easy-to-use single 20-pin connector. Figure 2.28 shows an example of an ATX system connector.

When the Pentium 4 processor was introduced, it required much more power than previous CPU models. Power measured in watts is a multiplicative function of voltage and current. To keep the voltage low meant that amperage would have to increase, but it wasn't feasible to supply such current from the power supply itself. Instead, it was decided to deliver 12V at lower amperage to a voltage regulator module (VRM) near the CPU. The higher current at a lower voltage was possible at that shorter distance from the CPU.

As a result of this shift, motherboard and power supply manufacturers needed to get this more varied power to the system board. The solution was the ATX12V 1.0 standard, which added two supplemental connectors. One was a single 6-pin auxiliary connector that supplied additional +3.3V and +5V leads and their grounds. The other was a 4-pin square mini-version of the ATX connector, referred to as a P4 (for the processor that first required them) connector, which supplied two +12V leads and their grounds. EPS12V uses an 8-pin version, called the processor power connector, which doubles the P4's function with four +12V leads and four grounds. Figure 2.29 illustrates the P4 connector. The 8-pin processor power connector is similar but has two rows of 4 and, despite its uncanny resemblance, is keyed differently from the 8-pin PCIe power connector to be discussed shortly.

PCIe devices require more power than PCI ones did. So, for ATX motherboards with PCIe slots, the 20-pin system connector proved inadequate. This led to the ATX12V 2.0 standard and the even higher-end EPS12V standard for servers. These specifications call for a 24-pin connector that adds further positive voltage leads directly to the system connector. The 24-pin connector looks like a larger version of the 20-pin connector. The corresponding pins of the 24-pin motherboard header are actually keyed to accept the 20-pin connector. Adapters are available if you find yourself with the wrong combination of motherboard and power supply. Some power supplies feature a 20-pin connector that snaps together with a separate 4-pin portion for flexibility, called a 20+4 connector, which can be seen in Figure 2.30. Otherwise, it will just have a 24-pin connector. The 6-pin auxiliary connector disappeared with the ATX12V 2.0 specification and was never part of the EPS12V standard.

ATX12V 2.1 introduced a different 6-pin connector, which was shaped a lot like the P4 connector (see Figure 2.31). This 6-pin connector was specifically designed to give additional dedicated power to the PCIe adapters that required it. It provided a 75W power source to such devices.

ATX12V 2.2 replaced the 75W 6-pin connector with a 150W 8-pin connector, as shown in Figure 2.32. The plastic bridge between the top two pins on the left side in the photo keeps installers from inserting the connector into the EPS12V processor power header but clears the notched connector of a PCIe adapter. The individual pin keying should avoid this issue, but a heavy-handed installer could defeat that. The bridge also keeps the connector from being inserted into a 6-pin PCIe header, which has identically keyed corresponding pins.

Proprietary Power Connectors

Although the internal peripheral devices have standard power connectors, manufacturers of computer systems sometimes take liberties with the power interface between the motherboard and power supply of their systems. It's uncommon but not unheard of. In some cases, the same voltages required by a standard ATX power connector are supplied using one or more proprietary connectors. This makes it virtually impossible to replace power supplies and motherboards with other units “off the shelf.” Manufacturers might do this to solve a design issue or simply to ensure repeat business.

SATA Power Connectors

SATA drives arrived on the market with their own power requirements in addition to their new data interfaces. (Refer back to Figure 2.9 to see the SATA data and power connectors.) You get the 15-pin SATA power connector, a variant of which is shown in Figure 2.33. The fully pinned connector is made up of three +3.3V, three +5V, and three +12V leads interleaved with two sets of three ground leads. Each of the five sets of three common pins is supplied by one of five single conductors coming from the power supply. When the optional 3.3V lead is supplied, it is standard to see it delivered on an orange conductor.

Note that in Figure 2.33, the first three pins are missing. These correspond to the 3.3V pins, which are not supplied by this connector. This configuration works fine and alludes to the SATA drives' ability to accept Molex connectors or adapters attached to Molex connectors, thus working without the optional 3.3V lead.

Multiple PSUs

It's almost unheard of to see two power supplies installed in a desktop computer. There's generally no need for such a setup and it would just be a waste of money. And for laptops and mobile devices, it's simply not an option. For servers, though, having a redundant power supply (RPS), meaning a second PSU installed in the system, might make sense. The sole reason to have two power supplies is in case one fails, the other can take over. The transition between the two is designed to be seamless and service will not be disrupted.

Based on its name and our description so far, it might seem as though this means installing two full-sized PSUs into a computer case. Given the limited amount inside a case, you can imagine how problematic this could be. Fortunately, though, PSU manufacturers make devices that have two identical PSUs in one enclosure. One such example is shown in Figure 2.36. The total device is designed to fit into ATX cases and is compliant with ATX12V and EPS12V standards. If one fails, the other automatically takes over. They are hot-swappable, so the failed unit can be replaced without powering the system down.

Although an RPS can help in the event of a PSU failure, it can't keep the system up and running if there is a power outage.

Battery Backup Systems

The second type of power redundancy is a battery backup system that the computer plugs into. This is commonly referred to as an uninterruptible power supply (UPS).

These devices can be as small as a brick, like the one shown in Figure 2.37, or as large as an entire server rack. Some just have a few indicator lights, whereas others have LCD displays that show status and menus and come with their own management software. The back of the UPS will have several power plugs. It might divide the plugs such that a few of them provide surge protection only, whereas others provide surge protection as well as backup power, as shown in Figure 2.38.

Inside the UPS are one or more batteries and fuses. Much like a surge suppressor, a UPS is designed to protect everything that's plugged into it from power surges. UPSs are also designed to protect against power sags and even power outages. Energy is stored in the batteries, and if the power fails, the batteries can power the computer for a period of time so that the administrator can then safely power it down. Many UPSs and operating systems will also work together to safely power down automatically a system that gets switched to UPS power. These types of devices may be overkill for Uncle Bob's machine at home, but they're critically important fixtures in server rooms.

The UPS should be checked periodically to make sure that its battery is operational. Most UPSs have a test button that you can press to simulate a power outage. You will find that batteries wear out over time, and you should replace the battery in the UPS every couple of years to keep the UPS dependable.

Monitors

Most display systems work the same way. First, the computer sends a signal to a device called the video adapter—an expansion board installed in an expansion bus slot or the equivalent circuitry integrated into the motherboard—telling it to display a particular graphic or character. The adapter then renders the character for the display—that is, it converts the single instruction into several instructions that tell the display device how to draw the graphic and sends the instructions to the display device based on the connection technology between the two. The primary differences after that are in the type of video adapter you are using (digital or analog) and the type of display (LCD, LED, IPS, and so forth).

Projection Systems

Another major category of display device is the video projection system, or projector. A portable projector can be thought of as a condensed video display with a lighting system that projects the image onto a screen or other flat surface for group viewing. Interactive whiteboards have become popular over the past decade to allow presenters to project an image onto the board as they use virtual markers to draw electronically on the displayed image. Remote participants can see the slide on their system as well as the markups made by the presenter. The presenter can see the same markups because the board transmits them to the computer to which the projector is attached, causing them to be displayed by the projector in real time.

To accommodate using portable units at variable distances from the projection surface, a focusing mechanism is included on the lens. Other adjustments, such as keystone, trapezoid, and pincushion, are provided through a menu system on many models as well as a way to rotate the image 180 degrees for ceiling-mount applications.

The key characteristics of projectors are resolution and brightness. Resolutions are similar to those of computer monitors. Brightness is measured in lumens. A lumen (lm) is a unit of measure for the total amount of visible light that the projector gives off, based solely on what the human eye can perceive and not on invisible wavelengths. Sometimes the brightness is even more of a selling point than the maximum resolution that the system supports because of the chosen environment in which it operates. For example, it takes a lot more to display a visible image in a well-lit office than it does in a darkened theater.

If you are able to completely control the lighting in the room where the projection system is used, producing little to no ambient light, a projector producing as little as 1,300 lumens is adequate in a home theater environment, while you would need one producing around 2,500 lumens in the office. However, if you can only get rid of most of the ambient light, such as by closing blinds and dimming overhead lights, the system should be able to produce 1,500 to 3,500 lumens in the home theater and 3,000 to 4,500 lumens in the office. If you have no control over a very well-lit area, you'll need 4,000 to 4,500 lumens in the home theater and 5,000 to 6,000 lumens in the business setting.

By way of comparison, a 60W standard light bulb produces about 800 lumens. Output is not linear, however, because a 100W light bulb produces over double, at 1,700 lm. Nevertheless, you couldn't get away with using a standard 100W incandescent bulb in a projector. The color production is not pure enough and constantly changes throughout its operation due to deposits of soot from the burning of its tungsten filament during the production of light. High-intensity discharge (HID) lamps, like the ones found in projection systems, do more with less by using a smaller electrical discharge to produce far more visible light. Expect to pay considerably more for projector bulbs than for standard bulbs of a comparable wattage.

Webcams

Years ago, owing to the continued growth of the Internet, video camera–only devices known as webcams started their climb in popularity. Today, with the prevalence of working from home and services like Zoom and Google Meet, it seems that everyone has been introduced to webcams.

Webcams make great security devices as well. Users can keep an eye on loved ones or property from anywhere that Internet access is available. Care must be taken, however, because the security that the webcam is intended to provide can backfire on the user if the webcam is not set up properly. Anyone who happens upon the web interface for the device can control its actions if there is no authentication enabled. Some webcams provide a light that illuminates when someone activates the camera. Nevertheless, it is possible to decouple the camera's operation and that of its light.

Nearly every laptop produced today has a webcam built into its bezel. An example is shown in Figure 3.8—this one has a light and two microphones built in next to it. If a system doesn't have a built-in camera, a webcam connects directly to the computer through an I/O interface, typically USB. Webcams that have built-in wired and wireless NIC interfaces for direct network attachment are prevalent as well. A webcam does not have any self-contained recording mechanism. Its sole purpose is to transfer its captured video directly to the host computer, usually for further transfer over the Internet—hence, the term web.

Keyboard

The keyboard is easily the most popular input device, so much so that it's more of a necessity. Very few users would even think of beginning a computing session without a working keyboard. Fewer still would even know how. The U.S. English keyboard places keys in the same orientation as the QWERTY typewriter keyboards, which were developed in the 1860s. Wired keyboards are almost always attached via USB. Wireless keyboards will often have a USB dongle that is attached to the computer, but they can also use Bluetooth.

Keyboards have also added separate number pads to the side and function keys (not to be confused with the common laptop key labeled Fn), placed in a row across the top of the keyboard above the numerical row. Key functionality can be modified by using one or more combinations of the Ctrl, Alt, Shift, and laptop Fn keys along with the normal QWERTY keys.

Technically speaking, the keys on a keyboard complete individual circuits when each one is pressed. The completion of each circuit leads to a unique scan code that is sent to the keyboard connector on the computer system. The computer uses a keyboard controller chip or function to interpret the code as the corresponding key sequence. The computer then decides what action to take based on the key sequence and what it means to the computer and the active application, including simply displaying the character printed on the key.

In addition to the layout for a standard keyboard, other keyboard layouts exist—some not nearly as popular. For example, without changing the order of the keys, an ergonomic keyboard is designed to feel more comfortable to users as they type. The typical human's hands do not rest with the fingers straight down. Ergonomic keyboards, therefore, should not place keys flat and along the same plane. To accomplish that goal, manufacturers split the keyboard down the middle, angling keys on each side downward from the center. Doing so fits the keys to the fingers of the hands when they are in a relaxed state. Figure 3.11 shows an example of an ergonomic keyboard. Even more exotic-looking ergonomic keyboards exist and may provide better relief for users who suffer from repetitive-use problems such as carpal tunnel.

Mouse

Although the computer mouse was born in the 1970s at Xerox's Palo Alto Research Center (PARC), it was in 1984 that Apple made the mouse an integral part of the personal computer with the introduction of the Macintosh. In its most basic form, the mouse is a hand-fitting device that uses some form of motion-detection mechanism to translate its own physical two-dimensional movement into onscreen cursor motion. Many variations of the mouse exist, including trackballs, tablets, touch pads, and pointing sticks. Figure 3.12 illustrates the most recognizable form of the mouse.

The motion-detection mechanism of the original Apple mouse was a simple ball that protruded from the bottom of the device so that when the bottom was placed against a flat surface that offered a slight amount of friction, the mouse would glide over the surface but the ball would roll, actuating two rollers that mapped the linear movement to a Cartesian plane and transmitted the results to the software interface. This method of motion detection has been replaced by optical receptors to catch LED light reflected from the surface the mouse is used on. Note that most optical mice will have problems working on a transparent glass surface because of the lack of reflectivity.

The mouse today can be wired to the computer system or connected wirelessly. A wired mouse typically uses a USB port, which also provides power. Wireless versions will have a USB dongle or connect via Bluetooth. They are powered with batteries, and the optical varieties deplete these batteries more quickly than their mechanical counterparts.

The final topic is one that is relevant for any mouse: buttons. The number of buttons that you need your mouse to have depends on the software interfaces you use. For the Macintosh, one button has always been sufficient, but for a Windows-based computer, at least two are recommended—hence, the term right-click. Today, the mouse is commonly found to have a wheel on top to aid in scrolling and other specialty movement. The wheel has even developed a click in many models, sort of an additional button underneath the wheel. Buttons on the side of the mouse that can be programmed for whatever the user desires are common today as well.

There are several variants on pointer devices, such as trackballs. A trackball is like an inverted mouse. Both devices place the buttons on the top, which is where your fingers will be. A mouse places its tracking mechanism on the bottom, requiring that you move the entire assembly as an analogue for how you want the cursor on the screen to move. In contrast, a trackball places the tracking mechanism, usually a ball that is about one inch in diameter, on the top with the buttons. You then have a device that need not be moved around on the desktop and can work in tight spaces and on surfaces that would be incompatible with the use of a mouse. The better trackballs place the ball and buttons in such a configuration that your hand rests ergonomically on the device, allowing effortless control of the onscreen cursor.

Universal Serial Bus

Universal Serial Bus (USB) cables are used to connect a wide variety of peripherals, such as keyboards, mice, digital cameras, printers, scanners, hard drives, and network cards, to computers. USB was designed by several companies, including Intel, Microsoft, and IBM, and is currently maintained by the USB Implementers Forum (USB-IF).

USB technology is fairly straightforward. Essentially, it is designed to be Plug and Play—just plug in the peripheral and it should work, provided that the software is installed to support it. Many standard devices have drivers that are built into the common operating systems or automatically downloaded during installation. More complex devices come with drivers to be installed before the component is connected.

USB host controllers can support up to 127 devices, which is accomplished through the use of a 7-bit identifier. The 128th identifier, the highest address, is used for broadcasting to all endpoints. Realistically speaking, you'll probably never get close to this maximum. Even if you wanted to try, you won't find any computers with 127 ports. Instead, you would plug in a device known as a USB hub (shown in Figure 3.14) into one of your computer's USB ports, which will give you several more USB ports from one original port. Understand that a hub counts as a device for addressing purposes. Hubs can be connected to each other, but interconnection of host controllers is not allowed; each one and its connected devices are isolated from other host controllers and their devices. As a result, USB ports are not considered networkable ports. Consult your system's documentation to find out if your USB ports operate on the same host controller.

Another nice feature of USB is that devices can draw their power from the USB cable, so you may not need to plug in a separate power cord. This isn't universally true, though, as some peripherals still require external power.

USB Cables and Connectors

In order to achieve the full speed of the specification that a device supports, the USB cable needs to meet that specification as well. In other words, USB 1.x cables cannot provide USB 2.0 and 3.x performance, and USB 2.0 cables cannot provide USB 3.x performance. Otherwise, the connected device will have to fall back to the maximum version supported by the cable. This is usually not an issue, except for the lost performance, but some high-performance devices will refuse to operate at reduced levels. Note that all specifications are capable of Low Speed, which is a 1.5 Mbps performance standard that has existed since the beginning of USB time.

Throughout most of its history, USB has relied on a small suite of standard connectors. The two broad classifications of connectors are designated Type-A and Type-B connectors, and there are micro and mini versions of each. A standard USB cable has some form of Type-A connector on the end that plugs into the computer or hub, and some form of Type-B or proprietary connector on the device end. Figure 3.15 shows five classic USB 1.x/2.0 cable connectors. From left to right, they are as follows:

• Micro-USB
• Mini-USB
• Type-B
• Type-A female
• Type-A male

Small form factor devices, including many smartphones and smaller digital cameras, use a micro-USB or mini-USB connector, unless the manufacturer has developed its own proprietary connector. Micro-USB connectors (and modified ones) are popular with many Android phone manufacturers.

In 2014, a new connector named USB Type-C (or simply USB-C) was developed. USB-C is designed to replace Type-A and Type-B, and, unlike its predecessors, it's reversible. That means no more flipping the connector over several times to figure out which way it connects. Type-C cables will also be able to provide more power to devices than classic cables were. Figure 3.16 shows a Type-C connector and a Type-A connector. You can see that while the Type-A connector is rectangular-shaped, the Type-C connector has rounded corners and looks more like an elongated oval.

USB was designed to be a short-distance technology. Because of this, USB cables are limited in length. USB 1.x and 2.0 can use cables up to 5 meters long, whereas USB 3.x can use cables up to 3 meters long. The maximum length of a USB4 cable is even shorter yet at 80 centimeters (0.8 meters). In addition, if you use hubs, you should never use more than five hubs between the system and any component.

Despite the seemingly locked-up logic of USB connectivity, it is occasionally necessary to alter the interface type at one end of a USB cable. For that reason, there are a variety of simple, passive converters on the market with a USB interface on one side and a USB or different interface on the other. Along with adapters that convert USB Type-A to USB Type-B, there are adapters that will convert a male connector to a female one. In addition, you can convert USB to a lot of other connector types, such as USB to Ethernet (shown in Figure 3.17), USB to SATA, USB to eSATA, USB to PS/2, USB to serial, and a variety of others.

Lightning

Introduced in 2012 with the iPhone 5, the Lightning connector is Apple's proprietary connector for iPhones and iPads. It's an 8-pin connector that replaced Apple's previous 30-pin dock connector. A standard Lightning cable has a USB Type-A connector on one end and the Lightning connector on the other, as shown in Figure 3.18. It's not keyed, meaning that you can put it in with either edge up.

Lightning cables support USB 2.0. You will find cables that are USB-C to Lightning, as well as various Lightning adapters, such as those to HDMI, DisplayPort, audio, and Lightning to female USB Type-A (so you can plug a USB device into an iPad or iPhone).

There are rumors that Apple may do away with the Lightning connector in a future iPhone release and instead use USB-C. After all, Apple has added USB-C ports to laptops and iPads, and USB-C is the port of the future. The same rumors have persisted since the iPhone 8 was released in 2017, and it seems that Apple has little reason to move away from its proprietary connector.

Thunderbolt

Where there's lightning, there's thunder, right? Bad joke attempts aside, in computer circles Lightning connectors don't have anything to do with Thunder(bolt). Thunderbolt, created in collaboration between Intel and Apple and released in 2011, combines PCI Express 2.0 x4 with the DisplayPort 1.x technology. While it's primarily used for video (to replace DisplayPort), the connection itself can support multiple types of peripherals, much like USB does.

Thunderbolt Cables and Connectors

The most common Thunderbolt cable is a copper, powered active cable extending as far as 3 meters, which was designed to be less expensive than an active version of a DisplayPort cable of the same length. There are also optical cables in the specification that can reach as far as 60 meters. Copper cables can provide power to attached devices, but optical cables can't.

Additionally, and as is the case with USB, Thunderbolt devices can be daisy-chained and connected via hubs. Daisy chains can extend six levels deep for each controller interface, and each interface can optionally drive a separate monitor, which should be placed alone on the controller's interface or at the end of a chain of components attached to the interface.

As noted in Table 3.3, Thunderbolt changed connectors between versions 2 and 3. Figure 3.19 shows two Thunderbolt 2 interfaces next to a USB port on an Apple MacBook Pro. Note the standard lightning-bolt insignia by the port. Despite its diminutive size, the Thunderbolt port has 20 pins around its connector bar, like its larger DisplayPort cousin. Of course, the functions of all the pins do not directly correspond between the two interface types, because Thunderbolt adds PCIe functionality.

Starting with Thunderbolt 3, the connector was changed to standard USB-C connectors, as shown in Figure 3.20. Notice that the lightning bolt icon remained the same.

Converters are available that connect Thunderbolt connectors to VGA, HDMI, and DVI monitors. Active converters that contain chips to perform the conversion are necessary in situations such as when the technology is not directly pin-compatible with Thunderbolt—as with VGA and DVI-A analog monitor inputs, for example. Active converters are only slightly more expensive than their passive counterparts but still only a fraction of the cost of Thunderbolt hubs. One other advantage of active connectors is that they can support resolutions of 4k (3840 × 2160) and higher.

Video Graphics Array Connector

The Video Graphics Array (VGA) connector was the de facto video standard for computers for years and is still in use today. First introduced in 1987 by IBM, it was quickly adopted by other PC manufacturers. The term VGA is often used interchangeably to refer to generic analog video, the 15-pin video connector, or a 640 × 480 screen resolution (even though the VGA standard can support much higher resolutions). Figure 3.23 shows a VGA port, as well as the male connector that plugs into the port. Nearly all VGA connectors are blue.

VGA technology is the only one on the objectives list that is purely analog. It has been superseded by newer digital standards, such as DVI, HDMI, and DisplayPort, and it was supposed to be phased out starting in 2013. A technology this widely used will be around for quite a while, though, and you'll still see it occasionally in the wild (or still in use).

Digital Visual Interface

The analog VGA standard ruled the roost for well over a decade but it had a lot of shortcomings. Digital video can be transmitted farther and at higher quality than analog, so development of digital video standards kicked off in earnest. The first commercially available one was a series of connectors known collectively as Digital Visual Interface (DVI) and was released in 1999.

At first glance, the DVI connector might look like a standard D-sub connector. On closer inspection, however, it begins to look somewhat different. For one thing, it has quite a few pins, and for another, the pins it has are asymmetrical in their placement on the connector. The DVI connector is usually white and about an inch long. Figure 3.24 shows what the connector looks like coming from the monitor.

There are three main categories of DVI connectors:

• DVI-A DVI-A is an analog-only connector. The source must produce analog output, and the monitor must understand analog input.
• DVI-D DVI-D is a digital-only connector. The source must produce digital output, and the monitor must understand digital input.
• DVI-I DVI-I is a combination analog/digital connector. The source and monitor must both support the same technology, but this cable works with either a digital or an analog signal.

The DVI-D and DVI-I connectors come in two varieties: single-link and dual-link. The dual-link options have more conductors—taking into account the six center conductors—than their single-link counterparts; therefore, the dual-link connectors accommodate higher speed and signal quality. The additional link can be used to increase screen resolution for devices that support it. Figure 3.25 illustrates the five types of connectors that the DVI standard specifies.

DVI-A and DVI-I analog quality is superior to that of VGA, but it's still analog, meaning that it is more susceptible to noise. However, the DVI analog signal will travel farther than the VGA signal before degrading beyond usability. Nevertheless, the DVI-A and VGA interfaces are pin-compatible, meaning that a simple passive DVI-to-VGA adapter, as shown in Figure 3.26, is all that is necessary to convert between the two. As you can see, the analog portion of the connector, if it exists, comprises the four separate color and sync pins and the horizontal blade that they surround, which happens to be the analog ground lead that acts as a ground and physical support mechanism even for DVI-D connectors.

It's important to note that DVI-I cables and interfaces are designed to interconnect two analog or two digital devices; they cannot convert between analog and digital. DVI cables must support a signal of at least 4.5 meters, but better cable assemblies, stronger transmitters, and active boosters result in signals extending over longer distances.

One thing to note about analog versus digital display technologies is that all graphics adapters and all monitors deal with digital information. It is only the connectors and cabling that can be made to support analog transmission. Before DVI and HDMI encoding technologies were developed, consumer digital video display connectors could not afford the space to accommodate the number of pins that would have been required to transmit 16 or more bits of color information per pixel. For this reason, the relatively few conductors of the inferior analog signaling in VGA were appealing.

High-Definition Multimedia Interface

High-Definition Multimedia Interface (HDMI) is an all-digital technology that advances the work of DVI to include the same dual-link resolutions using a standard HDMI cable but with higher motion-picture frame rates and digital audio right on the same connector. HDMI was introduced in 2002, which makes it seem kind of old in technology years, but it's a great, fast, reliable connector that will probably be around for several years to come. HDMI cabling also supports an optional Consumer Electronics Control (CEC) feature that allows transmission of signals from a remote-control unit to control multiple devices without separate cabling to carry infrared signals.

HDMI cables, known as Standard and High Speed, exist today in the consumer space. Standard cables are rated for 720p resolution as well as 1080i, but not 1080p. High Speed cables are capable of supporting not only 1080p, but also the newer 4k and 8k technologies. Figure 3.27 shows an HDMI cable and port.

In June 2006, revision 1.3 of the HDMI specification was released to support the bit rates necessary for HD DVD and Blu-ray Disc. This version also introduced support for “deep color,” or color depths of at least one billion colors, including 30-, 36-, and 48-bit color. However, not until version 1.4, which was released in May 2009, was the High Speed HDMI cable initially required.

With version 1.4 came HDMI capability for the controlling system—the television, for instance—to relay Ethernet frames between its connected components and the Internet, alleviating the need for each and every component to find its own access to the LAN for Internet access. Both Standard and High Speed cables are available with this Ethernet channel. Each device connected by such a cable must also support the HDMI Ethernet Channel specification, however.

Additional advances that were first seen in version 1.4 were 3D support, 4K resolution (but only at a 30 Hz refresh rate), an increased 120 Hz refresh rate for the 1080 resolutions, and an Audio Return Channel (ARC) for televisions with built-in tuners to send audio back to an A/V receiver without using a separate output cable. Version 1.4 also introduced the anti-vibration Type-E locking connector for the automotive-video industry and cables that can also withstand vibration as well as the hot/cold extremes that are common in the automotive world.

Version 2.0 of HDMI (2013) introduced no new cable requirements. In other words, the existing High Speed HDMI cable is fully capable of supporting all new version 2 enhancements. These enhancements include increasing the 4K refresh rate to 60 Hz, a 21:9 theatrical widescreen aspect ratio, and 32-channel audio. Note that 7.1 surround sound comprises only eight channels, supporting the more lifelike Rec. 2020 color space and multiple video and audio streams to the same output device for multiple users. Version 2.0a, released in 2015, primarily added high dynamic range (HDR) video, but it does not require any new cables or connectors.

The most recent version (as of this writing) is HDMI 2.1, released in November 2017. Version 2.1 specifies a new cable type called 48G, which provides for 48 Gbps bandwidth. 48G cables are backward compatible with older HDMI versions. You can also use older cables with 48G-capable devices, but you just won't get the full 48 Gbps bandwidth. HDMI 2.1 also provides for 120 Hz refresh rates for 4k, 8k, and 10k video, and it supports enhanced Audio Return Channel (eARC), which is needed for object-based audio formats, such as DTS:X and Dolby Atmos.

Even though the HDMI connector is not the same as the one used for DVI, the two technologies are electrically compatible. HDMI is compatible with DVI-D and DVI-I interfaces through proper adapters, but HDMI's audio and remote-control pass-through features are lost. Additionally, 3D video sources work only with HDMI. Figure 3.28 shows a DVI-to-HDMI adapter between DVI-D and the Type-A 19-pin HDMI interface. Compare the DVI-D interface in Figure 3.28 to the DVI-I interface in Figure 3.26, and note that the ground blade on the DVI-D connector is narrower than that of the DVI-A and DVI-I connectors. The DVI-D receptacle does not accept the other two plugs for this reason, as well as because the four analog pins around the blade have no sockets in the DVI-D receptacle.

Unlike DVI-D—and, by extension DVI-I—DVI-A and VGA devices cannot be driven passively by HDMI ports directly. An HDMI-to-VGA adapter must be active in nature, powered either externally or through the HDMI interface itself.

HDMI cables should meet the signal requirements of the latest specification. As a result, and as with DVI, the maximum cable length is somewhat variable. For HDMI, cable length depends heavily on the materials used to construct the cable. Passive cables tend to extend no farther than 15 meters, while adding electronics within the cable to create an active version results in lengths as long as 30 meters.

DisplayPort

DisplayPort is a royalty-free digital display interface from the Video Electronics Standards Association (VESA) that uses less power than other digital interfaces and VGA. Introduced in 2008, it's designed to replace VGA and DVI. To help ease the transition, it's backward compatible with both standards, using an adapter. In addition, an adapter allows HDMI and DVI voltages to be lowered to those required by DisplayPort because it is functionally similar to HDMI and DVI. DisplayPort cables can extend 3 meters, unless an active cable powers the run, in which case the cable can extend to 33 meters. DisplayPort is intended primarily for video, but, like HDMI, it can transmit audio and video simultaneously.

Figure 3.30 shows a DisplayPort port on a laptop as well as a connector. The DisplayPort connector latches itself to the receptacle with two tiny hooks. A push-button mechanism serves to release the hooks for removal of the connector from the receptacle. Note the beveled keying at the bottom-left corner of the port.

The DisplayPort standard also specifies a smaller connector, known as the Mini DisplayPort (MDP) connector. The MDP is electrically equivalent to the full-size DP connector and features a beveled keying structure, but it lacks the latching mechanism present in the DP connector. The MDP connector looks identical to a Thunderbolt 2 connector, which we covered in the “Peripheral Cables and Connectors” section earlier in this chapter.

Serial Advanced Technology Attachment

The most common hard drive connector used today is Serial Advanced Technology Attachment (SATA). Figure 3.31 shows SATA headers, which you have seen before, and a SATA cable. Note that the SATA cable is flat, and the connector is keyed to fit into the motherboard header in only one way. SATA data cables have a 7-pin connector. SATA power cables have 15 pins and are wider than the data connector.

The SATA we've discussed so far is internal, but there's an external version as well, appropriately named external SATA (eSATA). It uses the same technology, only in an external connection. The port at the bottom center of Figure 3.32 is eSATA. It entered the market in 2003, is mostly intended for hard drive use, and can support up to 15 devices on a single bus.

Table 3.4 shows some of the eSATA specifications.

You will commonly see the third generation of eSATA (and SATA) referred to as SATA 6 or SATA 6 Gb/s. This is because if they called it SATA 3, there would be confusion with the second generation, which had transfer speeds of 3.0 Gbps.

An interesting fact about eSATA is that the interface does not provide power, which is a big negative compared to its contemporary high-speed serial counterparts. To overcome this limitation, there is another eSATA port that you might see, called Power over eSATA, eSATA+, eSATAp, or eSATA/USB. It's essentially a combination eSATA and USB port. Since the port is a combination of two others, neither sanctioning body officially recognizes it (which is probably why there are so many names—other companies call it what they want to). Figure 3.33 shows this port.

You can see that this port is slightly different from the one in Figure 3.32, and it's also marked with a USB icon next to the eSATA one. On the market, you can purchase cables that go from this port to an eSATA device and provide it with power via the eSATAp port.

Parallel Advanced Technology Attachment

Prior to SATA, the most popular hard drive connector was Integrated Drive Electronics (IDE), which has now been renamed Parallel Advanced Technology Attachment (PATA). There is no difference between PATA and IDE, other than the name. Figure 3.34 shows PATA connectors on a motherboard next to a PATA cable. Refer back to Chapter 2, Figure 2.9, to see a direct comparison of SATA and PATA connectors on a hard drive.

PATA drives use a 40-pin flat data cable, and there are a few things to note about it. First, there is an off-colored stripe (often red, pink, or blue) along one edge of the cable to designate where pin 1 is. On a PATA drive, pin 1 is always on the edge nearest the power connector. The second thing to note is that there are three connectors—one for the motherboard and two for drives. PATA technology specifies that there can be two drives per cable, in a primary and secondary configuration. The primary drive will be attached to the other end of the cable, and the secondary, if connected, will use the middle connector. In addition, the drive itself may need to be configured for primary or secondary by using the jumper block on the drive. Most PATA drives will auto-configure their status based on their position on the cable, but if there is a conflict, they can be manually configured.

Power is supplied by a 4-pin power connector known as a Molex connector, shown in Figure 3.35. If you have a PATA drive and a SATA-supporting power supply (or vice versa), you can buy an adapter to convert the power to what you need. The same holds true for data connectors as well.

Small Computer System Interface

A fourth type of hard drive connector is called Small Computer System Interface (SCSI). The acronym is pronounced “scuzzy,” even though the original designer intended for it to be called “sexy.” The most common usage is for storage devices, but the SCSI standard can be used for other peripherals as well. You won't see many SCSI interfaces in home computers—it's more often found in servers, dedicated storage solutions, and high-end workstations.

Early versions of SCSI used a parallel bus interface called SCSI Parallel Interface (SPI). Starting in 2005, SPI was replaced by Serial Attached SCSI (SAS), which, as you may guess, is a serial bus. If you compare SCSI to other popular drive interfaces at the time, SCSI was generally faster but more expensive than its counterparts, such as IDE.

SCSI Parallel Interface

Although it's essentially obsolete now, you might find some details of SPI interesting. The first standard, ratified in 1986, was an 8-bit bus that provided for data transfers of 5 Mbps. Because it was an 8-bit bus, it could support up to seven devices. (The motherboard or expansion card header was the eighth.) Each device needed a unique ID from 0 to 7, and devices were attached in a daisy-chain fashion. A terminator (essentially a big resistor) needed to be attached to the end of the chain; otherwise, the devices wouldn't function.

In 1994, the 8-bit version was replaced by a 16-bit version that supported up to 15 devices and had a transfer speed of 320 Mbps. Compared to the 100 Mbps supported by IDE at the time, you can see why people wanted SCSI!

SPI had different connectors, depending on the standard; 50-pin, 68-pin, and 80-pin connectors were commonly used. Figure 3.36 shows two 50-pin Centronics connectors, which were common for many years. Figure 3.37 shows a terminator, with the top cover removed so that you can see the electronics.

Serial Attached SCSI

Of the newer SCSI implementations, the one you will most likely encounter is SAS. For example, as we mentioned in Chapter 2, most 15,000 rpm hard drives are SAS drives. From an architectural standpoint, SAS differs greatly from SPI, starting with the fact that it's serial, not parallel. What they do share is the use of the SCSI command architecture, which is a group of commands that can be sent from the controller to the device to make it do something, such as write or retrieve data.

A SAS system of hard drives works much like the SATA and PATA systems you've already learned about. There's the controller, the drive, and the cable that connects it. SAS uses its own terminology, though, and adds a component called an expander. Here are the four components of a SAS system:

• Initiator   Think of this as the controller. It sends commands to target devices and receives data back from them. These can be integrated or an add-on card. Each initiator can have a direct connection to 128 devices.
• Target   This is the device, typically a hard drive. It can also be multiple hard drives functioning in a RAID array.
• Service Delivery Subsystem   The service delivery subsystem transmits information between an initiator and a target. Often this is a cable, but it can also be a server backplane (where multiple devices connect).
• Expander   An expander is a device that allows for multiple initiators to be combined into one service delivery subsystem. Through the use of expanders, one initiator can support up to 16,256 devices.

Figure 3.38 shows a SAS cable and connector. It's slightly wider than a SATA power and data connector together. The other end of a cable such as this might have an identical SAS connector or a mini-SAS connector, or it might pigtail into four SATA or mini-SAS connectors.

Table 3.5 lists SAS standards and maximum throughput.

Standard	Year	Throughput
SAS-1	2005	3 Gbps
SAS-2	2009	6 Gbps
SAS-3	2013	12 Gbps
SAS-4	2017	22.5 Gbps

SAS offers the following advantages over SPI:

• No terminator is required.
• Up to 16,256 devices can be connected to a single system.
• Each SAS device has its own link to the controller, so there are no issues with contention (when multiple devices try to use the same link at the same time, causing interference).
• SAS provides faster data transfer speeds than SPI.
• SAS devices are compatible with SATA 2.0 and higher—SATA drives can be connected to SAS controllers.

With the invention of super-fast M.2 and NVMe hard drives, which you learned about in Chapter 2, it's hard to say what the future of SAS is. For example, SAS-5 (45 Gbps) has been under development since around 2018, but there is no official release date and there seems to be no impetus to get it to market. Most likely, SAS will continue to have a place in corporate environments with large-scale storage solutions, while the others will provide leading-edge speed for the workstation environment, particularly among laptops and smaller devices.

Daisy-Wheel Printers

The first type of impact printer to know about is the daisy-wheel printer. This is one of the oldest printing technologies in use. These impact printers contain a wheel (called the daisy wheel because it looks like a daisy) with raised letters and symbols on each “petal” (see Figure 4.1). When the printer needs to print a character, it sends a signal to the mechanism that contains the wheel. This mechanism is called the print head. The print head rotates the daisy wheel until the required character is in place. An electromechanical hammer (called a solenoid) then strikes the back of the petal containing the character. The character pushes up against an inked ribbon that ultimately strikes the paper, making the impression of the requested character.

Daisy-wheel printers were among the first types of impact printer developed. Their speed is rated by the number of characters per second (cps) they can print. The earliest printers could print only two to four characters per second. Aside from their poor speed, the main disadvantage of this type of printer is that it makes a lot of noise when printing—so much so, in fact, that special enclosures were developed to contain the noise. There is also no concept of using multiple fonts; the font is whatever the character on the wheel looks like.

The daisy-wheel printer has a few advantages, of course. First, because it is an impact printer, you can print on multipart forms (like carbonless receipts), assuming that they can be fed into the printer properly. Sometimes, you will hear this type of paper referred to as impact paper. Second, it is relatively inexpensive compared to the price of a laser printer of the same vintage. Finally, the print quality is easily readable; the level of quality was given a name: letter quality (LQ). Today, LQ might refer to quality that's better than an old-school typewriter (if you're familiar with them) but not up to inkjet standards.

Dot-Matrix Printers

The other type of impact printer to understand is the dot-matrix printer. These printers work in a manner similar to daisy-wheel printers, but instead of a spinning, character-imprinted wheel, the print head contains a row of pins (short, sturdy stalks of hard wire). These pins are triggered in patterns that form letters and numbers as the print head moves across the paper (see Figure 4.2).

The pins in the print head are wrapped with coils of wire to create a solenoid and are held in the rest position by a combination of a small magnet and a spring. To trigger a particular pin, the printer controller sends a signal to the print head, which energizes the wires around the appropriate print wire. This turns the print wire into an electromagnet, which repels the print pin, forcing it against the ink ribbon and making a dot on the paper. The arrangement of the dots in columns and rows creates the letters and numbers that you see on the page. Figure 4.2 illustrates this process.

The main disadvantage of dot-matrix printers is their image quality, which can be quite poor compared to the quality produced with a daisy wheel. Dot-matrix printers use patterns of dots to make letters and images, and the early dot-matrix printers used only nine pins to make those patterns. The output quality of such printers is referred to as draft quality—good mainly for providing your initial text to a correspondent or reviser. Each letter looked fuzzy because the dots were spaced as far as they could be and still be perceived as a letter or image. As more pins were crammed into the print head (17-pin and 24-pin models were eventually developed), the quality increased because the dots were closer together. Dot-matrix technology ultimately improved to the point that a letter printed on a dot-matrix printer was almost indistinguishable from daisy-wheel output. This level of quality is known as near letter quality (NLQ).

Dot-matrix printers are noisy, but the print wires and print head are covered by a plastic dust cover, making them quieter than daisy-wheel printers. They also use a more efficient printing technology, so the print speed is faster (typically starting around 72 cps). Some dot-matrix printers (like the Epson DFX series, which can run up to 1550 cps) can print close to a page per second! Finally, because dot-matrix printers are also impact printers, they can use multipart forms. Because of these advantages, dot-matrix printers quickly made daisy-wheel printers obsolete.

Parts of a Typical Inkjet Printer

Inkjet printers are simple devices. They contain very few parts (even fewer than dot-matrix printers) and, as such, are inexpensive to manufacture. It's common today to have a $40 to $50 inkjet printer with print quality that rivals that of basic laser printers.

The printer parts can be divided into the following categories:

• Print head/ink cartridge
• Head carriage, belt, and stepper motor
• Paper feed mechanism
• Control, interface, and power circuitry

Head Carriage, Belt, and Stepper Motor

Another major component of the inkjet printer is the head carriage and the associated parts that make it move. The print head carriage is the component of an inkjet printer that moves back and forth during printing. It contains the physical as well as electronic connections for the print head and (in some cases) the ink reservoir. Figure 4.3 shows an example of a head carriage. Note the clips that keep the ink cartridge in place and the electronic connections for the ink cartridge. These connections cause the nozzles to fire, and if they aren't kept clean, you may have printing problems.

The stepper motor and belt make the print head carriage move. A stepper motor is a precisely made electric motor that can move in the same very small increments each time it is activated. That way, it can move to the same position(s) time after time. The motor that makes the print head carriage move is also often called the carriage motor or carriage stepper motor. Figure 4.4 shows an example of a stepper motor.

In addition to the motor, a belt is placed around two small wheels or pulleys and attached to the print head carriage. This belt, called the carriage belt, is driven by the carriage motor and moves the print head back and forth across the page while it prints. To keep the print head carriage aligned and stable while it traverses the page, the carriage rests on a small metal stabilizer bar. Figure 4.5 shows the entire system—the stepper motor, carriage belt, stabilizer bar, and print head carriage.