|
Authorized HP Commercial Reseller
Microsoft Partner Company
For your custom business software solutions, Max PC
recommends:
| |
This section of our site was
created to provide information designed to help you sort through the myriad of
PC- and network-related topics, jargon, and specifications. The section will
always be a work in progress, so be sure to periodically check for the latest
updates and additions. If you have technical questions about topics not
addressed here, please feel free to inquire via e-mail or by using the
Feedback feature of this site.
Glossary
Want to be able to "talk the talk,"
or maybe just better understand the clerk at your local PC store who sounds like
he's speaking a foreign language when you ask a question? We have compiled a
glossary of common PC and network terminology to help you get up to speed.
Click on the button below to jump
to the glossary.
The
Numbers Game
Advertisements and literature for
PC and network equipment are rife with bandwidth and speed specifications,
memory size specifications, and storage capacity specifications, but what do
they mean in the real world?
The basic individually addressable
unit of computer storage is the byte. Strictly speaking, a byte is comprised of
8 bits (binary digits), but the computer requires two additional bits are
required for bookkeeping purposes, so a byte is generally considered a 10-bit
entity. While the main memory of early PC's was measured in terms of kilobytes
(thousands of bytes), memory capacity has grown exponentially over the years.
Memory quickly grew to megabytes (millions of bytes), and is now measured in
terms of gigabytes (billions of bytes). Similarly, secondary storage capability,
such as disk capacity has grown enormously over the years, and terabyte
(trillions of bits) hard drives are now available.
PC microprocessors have also
rapidly evolved, moving from 8-bit devices through 16-bit processors to 32-bit,
and now 64-bit units. Concurrently, modern microprocessor speeds have increased
into the gigahertz (billions of cycles per second) range. With each new
generation of processor, instruction sets have become more sophisticated, as
well, endowing newer processors with native functionality that older processors
simply didn't have.
But how fast is fast, how big is
big, and how much is enough? Let's take a look at the most commonly noted
parameters.
Memory
Each new generation of operating
systems and application programs typically imposes a greater demand on any PC's
underlying architecture, especially memory. As more features and functions are
added, programs grow in size, requiring more memory to support them. However,
memory comes in two distinct flavors: physical and virtual. The capacity of a
PC's physical memory is determined by the size of its installed memory modules,
and the capacity of its virtual memory is determined by the size of it Pagefile
or paging system, which is an area of the computer's hard drive where "pages" of
memory data are temporarily stored and retrieved as needed.
Modern PC's are multi-tasking
machines, which means they can - and do - run many different programs
simultaneously. But computers are not generally equipped with enough physical
memory to contain all of the programs running on them at any given time, so one
of the functions performed by a PC's operating system is memory management. The
operating system shuffles memory snapshots between physical memory and the
Pagefile system, bringing them into physical memory as they require processing
by the CPU and out to disk as they enter "wait" states.
Both physical and virtual memory
can be increased in size to accommodate increased computing demands. Most modern
operating systems will automatically adjust the size of virtual memory, but
paging memory is a painfully slow process compared to simply accessing the PC's
main physical memory (often called DRAM). Consequently, increasing physical
memory to reduce paging can often produce very significant improvements in the
overall speed of any PC.
Okay, how much DRAM is enough, and
is it possible to have too much? To support most contemporary operating systems
and applications, 1 GB (Gigabyte) of DRAM should be considered a bare minimum,
and some applications require considerably more than that to run efficiently. If
you've had your PC for a few years, it may not be equipped with even 1 GB of
memory, so adding more will certainly improve its performance. However, if your
operating system is, like most, a 32-bit system, it will be unable to access
more than 3 GB of physical memory. In other words, more is better, but only up
to a point. Installing 8 GB of memory in a PC running a 32-bit operating system
won't buy you any improvements that 4 GB won't buy you. If your PC is running
one of the newer 64-bit operating systems, then installing 8 or even 16 GB is an
option.
CPU
The microprocessor responsible for
executing all your operating system services and running your application
programs is called its central processing unit (CPU).
To support growing
computational demands, CPU speeds of have increased dramatically over the years,
and their architectures have become more elaborate. With respect to overall PC
speed, a faster is always better than a slower one. However, the improvement
realized by replacing a 2.5 GHz (Gigahertz) processor with a 3 GHz will be
relatively small compared to increasing physical memory size or switching to a
faster hard drive. The reason for this is that most modern PC's are I/O bound.
Simply put, this means the CPU spends far more time waiting for data and
commands to be loaded up than it does processing them.
If you're interested in eking out
the last bit of performance improvement from your current PC, replacing your CPU
with a faster compatible processor is certainly an option, but it may not be
very cost-effective. With each new generation, CPU's have become much more
sophisticated, and chances are you'll be unable to replace your current CPU with
a newer generation processor, as CPU socket configurations and voltage
requirements have also typically changed with each generation. To make a
genuinely significant processor improvement, you'll probably need a new main
board that will support a newer model processor in addition to a new CPU, and
probably new memory, as well. If your PC is equipped with a CPU older than a
Pentium IV generation processor, your most cost-effective option from a CPU
standpoint will most likely be a new PC.
Storage
Not all that ago, an 10 GB
(Gigabyte) hard drive was considered a large drive. Now, we have flash drives
and smart cards with higher capacities. As PC's have become more versatile, and
operating systems and application programs have grown in size, storage demands
have increased. Additionally, media files can be very large, further increasing
storage capacity requirements. Fortunately, with the advent of USB, Fire Wire
and SATA (serial ATA), the possibility of simply adding a large external data
drive to augment storage capacity has become a reality.
The amount of storage space
necessary to meet your particular needs will, of course, depend on the manner in
which you use your PC. A modest 20 GB drive may be perfectly adequate for a
machine dedicated to word processing and exchanging email, while a 500 GB drive
may be too small for a video production workstation. Only you can be the judge
of when you need a larger drive, but a good rule of thumb is that you should
consider migrating to a larger drive once your drive's available free space
falls below 25%. The reason for this is that your PC will run significantly more
slowly after you pass that point.
When considering a new drive, note
the specifications that effect its performance, as well as its capacity. Drive
data buffer sizes, spindle speeds, and seek times can vary widely among
available drives of the same size, and contribute to significant performance
differences among drives with the same storage capacity. Modern PC's are I/O
bound, making fast disk access extremely important.
Small Networks
Many home users and most small
businesses own multiple PC's. Those owners who wish to share files, Internet
connectivity, and peripheral equipment need networks to accomplish their goals.
A network may be established with Ethernet communication cables and standard
network equipment, wireless WiFi devices, or a combination of both. A
combination network topography, as depicted in the small network diagram at the
right, provides the greatest versatility and expandability.
Wireless routers generally provide
LAN (Local Area Network) Ethernet hub or switch functionality and are equipped
with a number of Ethernet ports for hardwired connections to network devices. In
the diagram at the right, the wireless router functions as the hub for a
hardwired desktop PC and network-enabled printer. The desktop PC accesses the
printer via an Ethernet connection through the router.
Another common router function is
to provide WAN (Wide Area Network) connectivity to the Internet through a cable
modem. For users with DSL (Digital Subscriber Line) Internet connections, a DSL
modem may be substituted for the cable modem in the diagram at the right.
Wireless routers also provide WiFi
network connectivity for WiFi-enabled devices. However, not all routers are
equipped with wireless functionality. If wireless networking capability is one
of your objectives, ensure that you select a wireless-enabled router.
The three PC's near the bottom of
the diagram at the right are all equipped with WiFi capabilities. This enables
these computers to share files and to share a single Internet connection through
the cable modem, via the router. They are also able to share files with the
desktop PC via its Ethernet connection to the router, as well as share the
network-enabled printer.
In the example to the right, the
shared printer is network-enabled. However, a printer that is not
network-enabled can still be shared as a network resource. In that situation,
the printer would simply be connected directly to a PC, and the PC configured to
share the device with the other PC's.
This simple example network is a
good starting point for most users, and can be easily expanded as one's
networking needs grow. Access points, bridges, switches, distance extenders,
media players, VoIP phones and a variety of other devices are available as
additions to this small network.
Ethernet Cables
Most
small hardwired networks are connected with twisted-pair cables having 8
individually insulated wires (arranged into 4 twisted pairs) bundled together
inside their jackets. The most common of these cables are called Category 5
(CAT-5) and Category 6 (CAT-6), depending on their bandwidth capabilities. CAT-5
cables are suitable for both 10 Mb/sec and 100 Mb/sec networks, while CAT-6
cables are required for Gigabit network applications.
Although CAT-5 and CAT-6 cables
have four pairs of wires, 10BASE-T and 100BASE-TX network cables require only
two pair to operate correctly: pins 1 and 2 are used for transmit (TX)
capability, and pins 3 and 6 for receive (RX) capability. Since 10BASE-T and
100BASE-TX need only two pairs, it is possible, but not standard, to run two
network connections (or a network connection and two phone lines) over a
twisted-pair Ethernet cable by using the normally unused pairs in these 10 and
100 Mb/s applications. However, this is not possible with 1000BASE-T, because
1000BASE-T requires all four pairs to operate.
All twisted-pair Ethernet cables
are produced with the same wire insulation colors, and the TIA/EIA-568B standard
allows for the choice of either of two specifications with respect to the
conventional RJ-45 connector pin designations of 10BASE-T and 100BASE-TX
Ethernet cables. These two specifications, T568A and T568B, differ only in that
they swap the positions of the two pairs used for transmitting and receiving
(TX/RX). Refer to the tables at the left for these two pin designation
specifications.
Ethernet standards enable the
majority of twisted-pair cables to be wired straight through, end-to-end (i.e.
pin 1 to pin 1, pin 2  to pin 2 and so etc.), but some cables need to be wired as
crossover cables (receive to transmit and transmit to receive). A crossover
cable is constructed by terminating one end of a twisted-pair network cable
according to the T568A specification, and the other end according to the T568B
specification. This swaps the transmit and receive wires end-to-end. The "tip"
and "ring" terms used in the T568 specifications are holdovers from older
communication technologies, and equate to the positive and negative portions of
the connections.
A 10BASE-T or 100BASE-TX network
node, such as a PC, which transmits on pins 1 and 2 to other network devices and
receives on pins 3 and 6 is referred to as a DTE (Data Terminal Equipment)
device. A network device, such as a hub, which transmits on pins 3 and 6
and receives on pins 1 and 2 is referred to as a DCE (Data
Communications/Carrier Equipment) device. A DTE device can be connected to a DCE
device with a straight-through cable. However, a crossover cable is generally
required to connect two DTE devices, such as two PC's, to one another, unless a
special port, referred to as an MDI-X port is available on one or the other
device. MDI-X ports employ internal (embedded) crossover wiring to enable a
straight-through connection between two devices of the same type. Hub and switch
ports with such internal crossovers are usually labeled with "uplink" or “X”
designations.
If two
devices being connected to one another both support the 1000BASE-T standard,
they will be able to communicate with one another, regardless of the cable being
used or how it is wired. Also, although an Ethernet crossover cable is generally
required to connect two computers directly together without a switch or hub,
many modern PC Ethernet host adapters are capable of automatically detecting
when another computer is connected with a straight-through cable, and will
automatically introduce the required crossover, when necessary. If neither of
the computers is equipped with such capability, then a crossover cable is
required.
PC
Drives
Many interface standards for
personal computer drives have appeared in the marketplace over the years, and
have passed into history as technology advanced and more robust standards
replaced them. The three most common standards still in use are Parallel ATA,
Serial ATA, and SCSI. These are detailed below.
PATA
Drives
Parallel ATA (PATA) drives have
been around for several years. They are now gradually being supplanted by Serial
ATA devices, but are still found in many PC's. PATA drives must be properly configured and connected to work correctly. When two
drives are connected to the PC with a single PATA ribbon cable, one must be
configured as device 0 (commonly referred to as "master" drive) and the other
must be configured as device 1 (the "slave" drive). This is necessary to enable
both drives to share the cable without conflict. The master drive is the drive
that usually appears first to the computer's BIOS and/or operating system, and
in an old BIOS (486 era and older), the drives are often referred to by the BIOS
as "C" for the master and "D" for the slave following the DOS naming convention
for the active primary partitions on each.
Configuring the mode that a drive
will use is often determined by a jumper on the drive, which must be manually
set to master or slave. If there is a single device on a PATA ribbon cable, it
should generally be configured as master. However, some hard drives - Western
Digital, in particular - have a separate setting called "single" to be used this
situation.
 The new PATA standard requires
color-coded connectors for easy identification by both installer and cable
maker, with all three connectors different colors from one another. Typically,
the connector connector for the master drive is black and located at one end of
the cable, the connector for the slave is gray and located near the middle of
the cable, and the connector to be attached to the motherboard is blue and
located at the opposite end of the cable from the master drive connector. The
motherboard connector of an 80-conductor ribbon cable has the socket for pin 34
connected to ground inside the connector but not attached to any conductor of
the cable. Since the old 40 conductor cables do not ground pin 34, the presence
of a ground connection indicates that an 80 conductor cable is being used. The
wire for pin 34 is attached normally on the other connectors and is not
grounded. Consequently, installing an 80-conductor cable backwards (with the
black connector on the system board, the blue connector on the master device,
and the gray connector on the slave will ground pin 34 of the master device and
connect host pin 34 through to pin 34 of the slave, resulting in unpredictable
behavior.
The
cable's
gray slave connector omits any connection to pin 28 but connects pin 34
normally, while the black end connector connects both pins 28 and 34 normally. A
drive configuration option called "cable select" has come into fairly widespread
use, and a PATA device set to "cable select" will automatically configure itself
as master or slave, according to its position on the ribbon cable. This is is
controlled by pin 28. Since the host (PC) adapter grounds this pin, and the pin
is left open on the slave connector, a PATA device will configure itself as the
master if it sees that the pin is grounded. If the device sees that pin 28 is
open, it configures itself as the slave device.
SATA Drives
|
Pin # |
Mating |
Function |
| |
1 |
3rd |
3.3 V |
|
2 |
3rd |
|
3 |
2nd |
| |
4 |
1st |
Ground |
|
5 |
2nd |
|
6 |
2nd |
| |
7 |
2nd |
5 V |
|
8 |
3rd |
|
9 |
3rd |
| |
10 |
2nd |
Ground |
| |
11 |
3rd |
Staggered spin-up/activity
(in supporting drives) |
| |
12 |
1st |
Ground |
| |
13 |
2nd |
12 V |
|
14 |
3rd |
|
15 |
3rd |
 |
|
15-pin SATA power connector. |
|
Pin # |
Function |
|
1 |
Ground |
|
2 |
A+ (Transmit) |
|
3 |
A− (Transmit) |
|
4 |
Ground |
|
5 |
B− (Receive) |
|
6 |
B+ (Receive) |
|
7 |
Ground |
|
- |
coding notch |
 |
|
7-pin SATA data cable. |
Newer PC's are typically equipped
with Serial ATA (SATA) drives. Many of these PC's are contain SATA controllers
on their motherboards. Others have SATA controllers installed in PCI or PCI
Express slots. Some SATA controllers are equipped with multiple ports for
connection to multiple devices. Also, port expanders or multipliers are
available to enable the connection of multiple devices to to a single SATA
controller port.
SATA was designed as a
point-to-point architecture, i.e. the connection between the controller and each
SATA device is direct. SATA drives use only 4 signal lines, so SATA cables are
more compact and less expansive to manufacture than PATA ribbon cables.
Additionally, the SATA interface standard supports hot-swapping, and an eSATA
connector is available for external SATA devices. SATA drives may also be
plugged into Serial Attached SCSI (SAS) controllers and communicate on the same
physical cable as native SAS disks, although SATA controllers cannot accommodate
SAS disks.
The SATA standard specifies a data
cable with seven conductors (3 grounds plus 4 data lines in two pairs) having
8-mm wide wafer connectors on each end. SATA cables can up to 1 meter in length,
and each connects one motherboard SATA socket to one device. Compared to PATA
ribbon cables, SATA cables are easier to install into confined spaces and
present fewer obstructions to cooling airflow within the PC case. However, SATA
connectors are more susceptible to accidental unplugging and breakage than PATA.
The motherboard SATA connector contains a plastic tab that can be broken if the
connector is bent, which may happen if the cable is pulled to one side. Because
the broken connector is on the motherboard, rather than the cable, it is not
easily replaced.
The SATA standard also specifies a new power connector. This
is a wafer-type connecter like the data cable, but its wider 15-pin design
prevents accidental misidentification and forced insertion of the wrong
connector into a SATA device's socket. SATA devices generally favor the SATA
power-connector over the older four-pin Molex connector found on most PATA
equipment. However, some SATA drives provide for attachment of the older 4-pin
Molex, as well as the SATA power connector.
SATA power connectors must contain more pins than the
traditional 4-pin connector for the following reasons:
-
3.3 V power is supplied in
addition to the traditional 5 V and 12 V sources.
-
Each voltage transmits
through three pins ganged together because the pins are so small that each,
by itself, cannot support the current required by some devices.
-
Five pins must be ganged
together to provide adequate grounding.
-
One of the three pins for
each voltage, provides for hot swapping. The ground pins and power pins 3,
7, and 13 are longer on the plug for the SATA device, so they will connect
first. A special hot-plug receptacle on the cable or backplane can connect
ground pins 4 and 12 first.
-
Pin 11 may optionally be used
to provide staggered spin-up functionality, activity indication, or no
function at all. Staggered spin-up may be implemented on drives to prevent
multiple units from spinning up simultaneously to avoid a potential
over-current situation. Alternatively, activity indication may implemented
to signal when the drive is busy.
Adaptors are available to convert 4-pin Molex connectors to
SATA wafer-style power connectors. However, because 4-pin Molex connectors do
not provide 3.3 V power, these adapters provide only 5 V and 12 V power, and
leave the 3.3 V lines unconnected, precluding the use of such adapters with
drives that require 3.3 V power. Fortunately, anticipating the use of these
adapters, many drive manufacturers have left the 3.3 V power lines unused by
their devices. However, a SATA device may be unable to implement the hot
swapping functionality mentioned above without 3.3 V power.
SCSI
Drives
The Small Computer System Interface
(SCSI) standard has been around since 1986, but continues to evolve and remains
a viable alternative to ATA for many applications. Although the SCSI standard
has traditionally been employed primarily for hard disks and tape drives, the
standard supports a wide range of other devices, including scanners and optical
drives. The SCSI standard defines command sets for specific peripheral device
types, but the inclusion of and "unknown" device type in the standard indicates
that it can, in theory, be used as an interface to almost any device.
SCSI is currently popular only on
high-performance workstations and servers. For example, RAID implementations on
servers almost always consist of SCSI hard disks, although a number of
manufacturers now offer SATA-based RAID systems as less expensive alternatives.
Although desktop computers and notebooks more typically use the ATA/IDE or the
newer SATA interfaces for hard disks, and USB, eSATA, and FireWire connections
for external devices, SCSI still has a place in the professional computing
world. Also, many other interfaces which do not implement the entire SCSI
standard still use SCSI command protocol.
In SCSI terminology, communication
takes place between an initiator and a target. The initiator sends a command to
the target which then responds. SCSI commands are sent in a Command Descriptor
Block (CDB). The CDB consists of a one byte operation code followed by five or
more bytes containing command-specific parameters. At the end of the command
sequence the target returns a Status Code byte which is usually 00H for success,
02H for an error (called a Check Condition), or 08H for busy. When the target
returns a Check Condition in response to a command, the initiator usually then
issues a SCSI Request Sense command in order to obtain a Key Code Qualifier (KCQ)
from the target. The Check Condition and Request Sense sequence involves a
special SCSI protocol called a Contingent Allegiance Condition.
SCSI has been available in a
variety of interfaces. The original, and still very common, is parallel SCSI,
now also called SPI, which uses a parallel electrical bus design. In 2008, the
Serial Attached SCSI (SAS) began to replace SPI. As the name implies, SAS
employs a serial design, but it retains other aspects of the SCSI technology.
The primary reason for the shift to serial interfaces is the clock skew issue of
high speed parallel interfaces, which makes the faster variants of parallel SCSI
susceptible to problems caused by cabling and termination, but in addition to
faster data rates, SAS also supports hot-swapping and improved fault isolation.
Another SCSI variant, iSCSI, uses TCP/IP as a transport mechanism.
On a parallel SCSI bus, each device
(e.g. host adapter or disk drive) is identified by a "SCSI ID" number. This is a
number in the range 0-7 on a narrow bus and in the range 0–15 on a wide bus.
Older controllers contained physical jumpers or switches to set the SCSI ID of
the initiator (host adapter). Modern host adapters (since about 1997), set their
SCSI ID's differently. The adapter generally contains a BIOS program that runs
when the computer boots, and this program provides menus enabling the user to
select the adapter's desired SCSI ID. Alternatively, the host adapter may come
with software that must be installed on the host computer to configure the SCSI
ID. The traditional SCSI ID for a host adapter is 7, because that ID is always
assigned the highest priority during bus arbitration, even on a 16-bit bus.
Setting the bootable (or first) hard disk to SCSI ID 0 is an accepted IT
community tradition. SCSI ID 2 is usually assigned to the floppy disk drive
while SCSI ID 3 is typically used for a CD-ROM drive.
Modern SCSI transport protocols
employ an automated process of ID discovery. SSA initiators "walk the loop" to
determine what devices are there and then assign each one a 7-bit "hop-count"
value. FC-AL initiators use the LIP (Loop Initialization Protocol) to
interrogate each device port for its WWN (World Wide Name). For iSCSI, because
of the unlimited scope of the (IP) network, the process is quite complicated.
These discovery processes occur at power-on/initialization time and also if the
bus topology changes later, for example if an new device is added.
|
